Seismic Analysis and Design of Hybrid Concrete Timber Structures With 2015 National Building Code of

SEISMIC ANALYSIS AND DESIGN OF HYBRID
CONCRETE TIMBER STRUCTURES WITH 2015

NATIONAL BUILDING CODE OF CANADA
by
Christopher Gregory Pieper
B.A.Sc., The University of British Columbia, 2014
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF APPLIED SCIENCE
in
THE COLLEGE OF GRADUATE STUDIES
(Civil Engineering)
THE UNIVERSITY OF BRITISH COLUMBIA
(Okanagan)
October 2018
© Christopher Gregory Pieper, 2018

The following individuals certify that they have read, and recommend to the College of Graduate
Studies for acceptance, a thesis/dissertation entitled:
SEISMIC ANALYSIS AND DESIGN OF HYBRID CONCRETE TIMBER

STRUCTURES WITH 2015 NATIONAL BUILDING CODE OF CANADA
submitted by Christopher Gregory Pieper in partial fulfillment of the requirements of
the degree of Master of Applied Science.
Dr. Solomon Tesfamariam, School of Engineering (Civil)

Supervisor
Dr. Shahria Alam, School of Engineering (Civil)

Supervisory Committee Member
Dr. Kasun Hewage, School of Engineering (Civil)

Supervisory Committee Member
Dr. Thomas Johnson, School of Engineering (Electrical)

University Examiner
ii
Abstract
The increasing use of timber for mid to high-rise structures fosters a need for high quality
numerical modelling and modelling input parameters for timber and timber hybrid structural
systems. Concrete is still a primary building material for mid to high-rise structures; however,
designers seek to utilize timber, concrete, and steel to develop innovative structural systems.
Experimental timber connection data was used to validate Pivot hysteretic connection models.
Results of validated tension-bolted timber connection parameters for timber column to foundation,
timber beam to timber column, and timber beam to concrete core are provided for use with
numerical models.
Experimental frame and connection data was used to validate connection and two-dimensional
timber frame models. Similarly, results of hysteretic connection models for parametrized timber
frame beam to column and column to base lagscrewbolt type connections are provided; as well,
the numerical frame models developed demonstrate good agreeance with the experimental data.
Experimental results from full scale shake table testing were used to validate a two-storey concrete
core with eccentric timber frame three-dimensional numerical structural model. With limitations
noted, results somewhat predict response of the hybrid structure to cyclic ground motion. Critical
parameters of hysteretic connection performance, damping, diaphragm stiffness, and core wall
cracking are explored.
By extension of the experimental validations, a 12-storey hybrid concrete core timber frame
numerical model was developed and analysed with equivalent static force procedure, modal
response spectrum, linear response history and non-linear response history analysis. The selection
and scaling of ground motions was completed following Method A from the newly released 2015
NBCC guideline; the selected scenario-specific ground motions, required processing, and
individual and suite scaling factors are provided for three suites of five bi-directional ground
motion pairs. Following Part 4, Division B, Section 4.1.8 of the 2015 National Building Code of
Canada, the concrete core was designed for strength and drift using ETABS; further, the tension-
bolted timber connections were able to sustain anticipated seismic shaking but not to contribute to
the seismic force resisting system.
iii
Lay Summary
Particularly in moderate to high seismic regions, earthquake engineering is required to ensure

structures will withstand earthquake shaking without collapse. Seismic structural design has
evolved significantly in the past 50 years such that desirable ductile (non-collapsing) failure modes
under significantly large earthquakes can be achieved. Even more recently, timber is being viewed
as a feasible building material for mid to high-rise structures; therefore, seismic timber and
particularly timber hybrid (mixed timber-concrete and/or timber-steel) engineering has become an
important topic for both research and professional practise in the analysis and design of structures.
One, two and three-dimensional numerical (computer) models of structures are developed.
Experimental data is used to validate each model and results are provided for general use. Finally,
experimental validations are extended to linear and non-linear analysis and design aspects of a 12-
storey hybrid concrete timber structure following the 2015 national building code of Canada.
iv
Preface
This thesis is original, unpublished work by the author under the supervision of Dr. Solomon
Tesfamariam. Experimental results were provided by Dr. Tesfamariam to complete each of the
experimental validations.
v
Table of Contents
Abstract .......................................................................................................................................... iii
Lay Summary ................................................................................................................................. iv
Preface............................................................................................................................................. v
Table of Contents ........................................................................................................................... vi
List of Tables .................................................................................................................................. x
List of Figures ................................................................................................................................ xi
Acknowledgements ...................................................................................................................... xiv
Dedication ..................................................................................................................................... xv
1 Introduction ............................................................................................................................. 1
1.1 Objectives ......................................................................................................................... 2
1.2 Scope ................................................................................................................................ 2
1.3 Motivations....................................................................................................................... 4
1.4 Organization ..................................................................................................................... 6
2 Literature Review .................................................................................................................... 7

2.1 Structural materials .......................................................................................................... 7
2.1.1 Reinforced concrete .................................................................................................. 7
2.1.2 Timber ....................................................................................................................... 8
2.2 Structural components .................................................................................................... 10
2.2.1 Foundations and anchorage..................................................................................... 10
2.2.2 Timber frames ......................................................................................................... 11
2.2.3 Timber floor diaphragms ........................................................................................ 12
2.2.4 Shear walls .............................................................................................................. 14
2.2.5 Coupled shear walls ................................................................................................ 15
2.2.6 Non-components – components not part of the SFRS ............................................ 16
vi
2.3 Seismic design considerations........................................................................................ 16
2.3.1 Ductility and overstrength....................................................................................... 16
2.3.2 Structural irregularity .............................................................................................. 17
2.3.3 Torsion .................................................................................................................... 18
2.3.4 Stability ................................................................................................................... 19
2.3.5 Seismic weight ........................................................................................................ 21
2.3.6 Damping .................................................................................................................. 21
2.3.7 Hysteresis ................................................................................................................ 22
2.3.8 Capacity design ....................................................................................................... 24
2.4 Code procedures and analysis methods .......................................................................... 25
2.4.1 Equivalent static force procedure............................................................................ 26
2.4.2 Modal response spectrum analysis.......................................................................... 26
2.4.3 Linear response history analysis ............................................................................. 27
2.4.4 Non-linear response history analysis ...................................................................... 28
2.4.5 Ground motion selection and calibration ................................................................ 29
3 Methodology.......................................................................................................................... 33
3.1 Research planning .......................................................................................................... 34
3.2 Examples chosen ............................................................................................................ 34
3.3 Data collection................................................................................................................ 35
3.4 Importance and limitations ............................................................................................. 35
3.5 Proposed analysis ........................................................................................................... 35
3.6 Summary ........................................................................................................................ 36
4 Case Details ........................................................................................................................... 37

4.1 Experimental validations ................................................................................................ 37
4.1.1 One-dimensional study ........................................................................................... 37
4.1.2 Two-dimensional study........................................................................................... 39
4.1.3 Three-dimensional study......................................................................................... 41
4.2 Extended study ............................................................................................................... 43
4.2.1 Case details from 12-storey structural systems ....................................................... 44
4.2.2 Case details for extended study initial analysis model ........................................... 48
vii
5 Procedures ............................................................................................................................. 50
5.2 Extended study ............................................................................................................... 58
5.2.1 Initial design steps................................................................................................... 58
5.2.2 Equivalent static force procedure............................................................................ 67
5.2.3 Modal response spectrum analysis.......................................................................... 73
5.2.4 Selection and scaling of ground motions ................................................................ 76
5.2.5 Linear response history analysis ............................................................................. 86
5.2.6 Non-linear response history analysis ...................................................................... 88
6 Results ................................................................................................................................... 90
6.2 Extended study ............................................................................................................... 96
6.2.1 Equivalent static force............................................................................................. 96
6.2.2 Response spectrum................................................................................................ 100
6.2.3 Selection and scaling of ground motions .............................................................. 101
6.2.4 Linear response history ......................................................................................... 102
6.2.5 Non-linear response history .................................................................................. 105
7 Conclusions ......................................................................................................................... 109

7.1 Experimental validations .............................................................................................. 109
7.1.1 One-dimensional study ......................................................................................... 109
7.1.2 Two-dimensional study......................................................................................... 111
7.1.3 Three-dimensional study....................................................................................... 112
7.2 Extended study ............................................................................................................. 113
viii
References ................................................................................................................................... 117
Appendices .................................................................................................................................. 128

Appendix A – Additional information for selection of ground motions ................................. 128
ix
List of Tables
Table 2-1 Analysis methods prescribed in the 2015 national building code of Canada .............. 26
Table 4-1 Structural loads and deflection criteria ........................................................................ 49
Table 5-1 2015 National building code interpolated seismic hazard values ................................ 59
Table 5-2 Geotechnical parameters provided for the site ............................................................ 59
Table 5-3 NBCC (2015) Table 4.1.8.4.-B ................................................................................... 61
Table 5-4 Structural loads and deflection criteria ........................................................................ 63
Table 5-5 Timber component sizes .............................................................................................. 64
Table 5-6 Preliminary seismic weights used for analysis ............................................................ 66
Table 5-7 Eccentricities and total static torsional moment .......................................................... 69
Table 5-8 Change in key parameters due to model update .......................................................... 72
Table 5-9 Modal response spectrum elastic base shear and fundamental period......................... 73
Table 5-10 Participating mass ratio's ........................................................................................... 78
Table 5-11 Target scenario's ........................................................................................................ 80
Table 5-12 Scenario-specific period range of interest ................................................................. 80
Table 6-1 Pivot connection parameters for tension bolt moment resisting connections .............. 92
Table 6-2 Pivot parameters for lagscrewbolts connections .......................................................... 93
Table 6-3 Final selected suites of ground motions ..................................................................... 101
Table 6-4 Member design forces from preliminary ground motions ......................................... 105
x
List of Figures
Figure 1-1 12-Storey concrete timber structure ............................................................................. 1
Figure 2-1 Hysteretic connection models .................................................................................... 23
Figure 4-1 Experimental apparatus for connection testing .......................................................... 37
Figure 4-2 Load protocol ............................................................................................................. 38
Figure 4-3 Connections ............................................................................................................... 39
Figure 4-4 Portal frame study ...................................................................................................... 39
Figure 4-5 Portal frame lagscrewbolt connections ...................................................................... 40
Figure 4-6 General arrangement of concrete-timber hybrid ........................................................ 41
Figure 4-7 Three-dimensional two-storey concrete-timber structure .......................................... 42
Figure 4-8 Concrete core section ................................................................................................. 43
Figure 4-9 General arrangement plans ......................................................................................... 45
Figure 4-10 General arrangement plan of 12-storey FFTT timber system .................................. 46
Figure 4-11 Core walls ................................................................................................................. 47
Figure 4-12 Preliminary load criteria ........................................................................................... 48
Figure 4-13 Preliminary structural model .................................................................................... 49
Figure 5-1 One-dimensional model with hysteretic spring .......................................................... 50
Figure 5-2 Two-dimensional non-linear portal frame model....................................................... 54
Figure 5-3 Hysteretic matching of column base in frame model ................................................. 55
Figure 5-4 Three-dimensional model linear concrete core and non-linear timber connections .. 56
xi
Figure 5-5 Parametric study of modal damping parameter.......................................................... 57
Figure 5-6 Design spectral response accelerations ...................................................................... 62
Figure 5-7 Preliminary shear and overturning results .................................................................. 69
Figure 5-8 Varying eccentricities and inherent torsion ................................................................ 75
Figure 5-9 Scaling of elastic forces.............................................................................................. 75
Figure 5-10 Scaling of drifts considering torsion ........................................................................ 76
Figure 5-11 Response spectrum for selection and scaling of ground motions ............................ 77
Figure 5-12 Period range of interest ............................................................................................ 78
Figure 5-13 Deaggregated hazard ................................................................................................ 79
Figure 5-14 Crustal suite .............................................................................................................. 83
Figure 5-15 In-slab suite .............................................................................................................. 84
Figure 5-16 Interface suite ........................................................................................................... 85
Figure 5-17 Governing maximum drift orthogonal pairs with suite average .............................. 87
Figure 5-18 General arrangement ................................................................................................ 89
Figure 6-1 Calibrated Pivot hysteretic connection models .......................................................... 91
Figure 6-2 Connection static pushover curves matched to experimental data ............................. 94
Figure 6-3 Frame static pushover curves matched to experimental data ..................................... 94
Figure 6-4 Best fit of maximum floor acceleration...................................................................... 95
Figure 6-5 Best fit of maximum floor displacement .................................................................... 96
Figure 6-6 Equivalent static drift ................................................................................................. 97
Figure 6-7 Equivalent static storey shear ..................................................................................... 97

xii
Figure 6-8 Equivalent static overturning moment ....................................................................... 97
Figure 6-9 Concrete core.............................................................................................................. 99
Figure 6-10 Response spectrum drift ......................................................................................... 100
Figure 6-11 Response spectrum storey shear ............................................................................. 100
Figure 6-12 Response spectrum overturning moment ............................................................... 100
Figure 6-13 Scaled mean spectra ............................................................................................... 101
Figure 6-14 Linear response history suite average maximum drift ........................................... 102
Figure 6-15 Linear response history suite average maximum storey shear ............................... 103
Figure 6-16 Linear response history suite average overturning moment .................................. 104
Figure 6-17 Non-linear response history suite average maximum drift .................................... 106
Figure 6-18 Non-linear response history suite average maximum storey shear ........................ 107
Figure 6-19 Non-linear response history suite average maximum overturning moment .......... 108
xiii
Acknowledgements
I would like to acknowledge my supervisor Dr. Solomon Tesfamariam, P.Eng. for providing
direction, knowledge, support, and patience throughout my graduate studies. I would also like to
thank Mr. Tom Guenther, P.Eng. of Okanagan College and TWG Structural Consultant for guiding
me to move forward with a bachelors program the opportunity to practise structural engineering
throughout my college and bachelor studies; Dr. Rafiqul Haque, EIT (formerly of University of
British Columbia) for his time in sharing his knowledge of finite element modelling techniques
and seismic engineering which allowed me to progress through challenging aspects of the
experimental validations; Mr. Tim Dunne, P.Eng. and Mr. Doug Clough, P.Eng. of Omega
Engineering SA LLP for providing financial support and for the opportunity to grow and practise
professionally during my studies; Dr. Stephen Halchuk of Natural Resources Canada for providing
deaggregation plots and taking time to address my questions related to seismicity; Dr. Robert
Tremblay of Polytechnique Montreal for his time in providing reference material and dialogue
related to the selection and scaling of ground motions; The collaborators with Dr. Tesfamariam
from the three experimental validation studies carried out in this thesis from which the
experimental data was made available; NSERC and the UBC Work Study program for funding
which facilitated my most preliminary studies of earthquake engineering and hybrid structures;
and, the numerous friends and colleagues who have helped to develop my structural engineering
abilities.
I would also like to acknowledge my family for their endless support during this thesis: My mom
and dad who have devoted many hours to my family while I studied; my sister who first studied
engineering and helped me believe I could as well; my loving children who step-up, provide
balance and inspire me; my wife who continually encourages and whole-heartedly believes in me;
my in-laws who help with the family while I focus; my grandfather who advocated for me to study
engineering.
Without all of their support, this would not have been possible.
xiv
Dedication
In commemoration of their tireless support, this thesis is dedicated to my family.
xv
1 Introduction
Modern building codes require numerical (computer) models for seismic design of mid to high-
rise structures. The validity of such models is critical to gaining expected performance of
constructed structures. Traditionally, concrete and steel structures dominated the mid to high-rise
landscape; however, timber is increasingly being included. With the relatively new inclusion of
timber in mid to high-rise structures, there is a need to develop accurate timber connection models
and analysis techniques to provide both economical and safe designs. With high seismic demand
in southwestern British Columbia, an abundance of timber as a construction material and a
historical precedent of reinforced concrete (concrete) being used for seismic force resisting (core)
systems, seismic structural design of hybrid concrete timber structures has become an important
topic for structural engineers practising in this field.
Figure 1-1 12-Storey concrete timber structure
1
1.1 Objectives
The objectives of this thesis are to:
1. Validate numerical hysteretic connection models, for timber and hybrid concrete timber
structures under reverse cyclic loading, with experimental results.
2. Validate non-linear numerical frame models, which include calibrated hysteretic connection
models, under non-linear static pushover analysis and reverse cyclic loading, with experimental
results.
3. Validate a numerical 3D hybrid structure model, which includes calibrated hysteretic connection
models, and variable damping, semi-rigid diaphragm, and cracked core stiffness parameters, to
experimental results.
4. Implement the national building code of Canada (2015) to an innovative hybrid concrete timber
structure numerical model to determine the effects of the timber frame semi-rigidities on the
structural system performance and design criteria
5. Implement calibrated connection models and damping into a code-based study of a 12-storey
hybrid concrete timber structure.
6. Implement code procedures for all four analysis types; equivalent static, response spectrum,
linear and non-linear response history.
7. Implement the new guidelines for selection and scaling of ground motions.
1.2 Scope
Seismic structural analysis and design aspects are presented for a twelve-storey concrete-core
timber-frame structure (see Figure 1-1) based on geometric plan layout and storey heights from
comparable structures (MGB Architecture, and Equilibrium Consulting, 2012).
2
Numerical models of hysteretic connections, frame assemblies and a three-dimensional structure
are fitted and validated with experimental data. Also using experimental results, structural
modelling techniques are developed and validated connection parameters are presented. Aspects
of seismic design are implemented in accordance with the 2015 national building code of Canada.
Recently, shake table tests were carried out and results obtained for a hybrid concrete-timber
structure (Isoda, Kawai, Koshihara, Araki, & Tesfamariam, 2016). Presently, the structure does
not appear to have been numerically modelled. Similarly, portal frame and timber lagscrewbolt
connection testing has been carried out (Mori, Nakatani, & Tesfamariam, 2015) and timber
tension-bolt connection testing has been carried out (Isoda et al., 2016; Isoda & Tesfamariam,
2016). Overall finite element models and their specific connection hysteresis models can be
calibrated to experimental data which improves both prediction of structural behaviour and overall
structural designs.
Experimental results were used to develop the baseline for an extended numerical study of a 12-
storey concrete-timber structure.
The experimental validations conducted using ETABS were:
1) One-dimensional study – Pivot connection model parameters are validated for timber and
timber-concrete tension-bolt connections using experimental results from reverse-cyclic
load tests presented in Isoda and Tesfamariam (2016)
2) Two-dimensional study – Timber portal frames and their connections are validated using
experimental results from monotonic static pushover and reverse-cyclic load tests (Mori et
al., 2015)
3) Three-dimensional study – A two-storey hybrid concrete-timber structure utilizing the
previously fitted tension-bolt connections is validated under lateral ground motion
(acceleration) input using available shake table results (Isoda et al., 2016)
The extended study consists of the selected seismic analysis and design procedures for a 12-storey
hybrid concrete timber structure following the national building code of Canada (2015) and
implementing each of the four prescribed analysis procedures. Further, the concrete core primary
seismic force resisting system is designed and influence of semi-rigid timber frames is determined.
3
A focus on code compliance and proper analysis steps is provided. In the equivalent static force
procedure and response spectrum analysis, consideration of stability, torsion, regularity and
capacity design are considered. Then, the primary seismic force resisting system is designed for
strength and interstorey drift limit. For both the linear and non-linear response history analysis, the
new (2015) NRC guidelines are followed to select and scale the ground motions for a three-
dimensional simultaneous bi-directional analysis. Results of the linear response history analysis
includes bi-directional pairs of major and minor demand for overturning as results from application
of 100:30 rule. Results from both linear and non-linear response history analysis are averaged by
scenario-specific suite such that design response can be selected based on critical suite(s) at each
location.
1.3 Motivations
The primary motivations for this thesis were as follows:
1. Experimental validations of structures
Accurate analysis models are critical to enable creation of designs that are both efficient and safe
for structures subjected to seismic loading. Experimental validations require that analysis
modelling techniques be refined and provide a check for modelling techniques. With experimental
validations not typically available in professional practice, the experimental validations provide
an invaluable baseline for numerical modelling of structures.
2. Experimental validations of timber connections
As use of timber has increased in mid to high-rise buildings, various innovative timber connections
are being created. Many timber connections are not yet codified and therefore designers must rely
on experimental testing and calibration of test data to connection models - which is often not
feasible to complete on a project specific basis. By providing connection values for various timber
connection models, implementation will be facilitated.
3. Building code provisions and design methodologies
In professional practice, building codes provide structural design requirements. Aside from current
and emerging research, the building code generally provides up to date and current state of the art
4
thinking. The 2015 national building code of Canada (NBCC) and supporting 2015 structural
commentaries, including the new guidelines for the selection and scaling of ground motions, are
likely to be enacted under a provincial version in British Columbia very soon; therefore, relevant
building code provisions must be well understood.
4. New guidelines for selection and scaling of ground motions
While ground motions are typically a geotechnical engineers’ responsibility, it is helpful for a
structural engineer to have a good understanding of the geotechnical aspects that impact their work.
The selection and scaling of historical ground motion records is currently the preferred method in
the NBCC (2015) for load input when conducting the increasingly required non-linear analysis of
structures.
5. Design of concrete core
Design of concrete core structures is well documented. For this prominent type of hybrid structure
(concrete-core timber-frame) to be designed, it is critical for designers to both be able to design
the core and determine the influence of the interacting timber components of the primary structural
system.
6. Market demand
Climate change science has caused governments to increasingly demand sustainable structures.
Further, global trends of urban densification promote construction of mid to high-rise structures.
Governments also aim for economic diversification; in Canada, this includes fostering the use of
timber and value-added timber products. Development of the timber and hybrid timber structural
engineering practise will assist in meeting the emerging market demand.
7. Professional responsibilities
It is incumbent on the professional engineer to maintain current knowledge in the field and promote
advances in their area of practice. Hybrid structures represent advances in building technology.
Non-linear and code complaint analysis and design is also important as designers must safely
design economical, sustainable and often intriguing structures.
5
1.4 Organization
In Chapter two, the literature review is presented. First, structural materials are discussed. Second,
structural components are explored. Third, seismic design considerations are brought to light.
Finally, code procedures for analysis and design are reviewed.
In Chapter three, the methodology is presented. A recap of the literature review findings, rational
for the experimental validations, and need to proceed to the extended study is explained.
In Chapter four, the case details of each study conducted in the thesis are presented. First, the
experimental validations are presented including one, two and three-dimensional studies. Next, the
details of the extended study are presented including details of the reference structures and
preliminary structural configuration adopted for analysis and design of the 12-storey structure.
In Chapter five, the procedures followed throughout the studies are presented. Each of the one,
two, and three-dimensional studies procedures are provided. Next, the procedures followed for
seismic analysis and design of the 12-storey structure according to the national building code of
Canada (2015) are provided.
In Chapter six, the results are provided. The experimental validation results provide timber and
timber hybrid tension-bolt and lagscrewbolt hysteretic Pivot connection parameters. Results of
structural model validation are provided for the three-dimensional two-storey hybrid structure.
Next, results from the numerical extension to analysis and design of a 12-storey structure are
provided. The results come from each of the four prescribed analysis methods of the NBCC (2015)
as well as the selection and scaling of ground motions.
In Chapter seven, the conclusions are presented. More precisely, this chapter includes conclusions,
recommendations, limitations, and future work. The chapter first looks at the experimental
validations then discusses the extended study.
6
2 Literature Review
This Chapter reviews 1) structural materials, 2) structural components, 3) seismic design

considerations, then 4) building code procedures relevant to seismic analysis and design of hybrid
concrete timber structures. This review also develops the rational for subsequent analysis and
design decisions of the extended study.
2.1 Structural materials
Concrete, timber, and steel are typical materials that form structural systems. Steel is not
investigated here except as reinforcing (rebar) in concrete and as connection material for timber.
Each material has unique material design standards (CSA - CSA Standards, 2014a; CSA - CSA
Standards, 2014b; CSA - CSA Standards, 2014c; Popovski et al., 2014). Reinforced concrete and
timber can be used together to form a hybrid component, system or structure. (Abeysekera, 2015;
Isoda et al., 2016; Schneider et al., 2015). A unique type of hybrid concrete timber system is a
concrete timber composite; research of such composites is ongoing (Dias, 2012; Dias, Van de
Kuilen, Lopes, & Cruz, 2007; Hehl, Tannert, Meena, & Vallee, 2014; Khorsandnia, 2015; Yeoh,
2011). Current research aims to identify promising hybrid systems (Abeysekera, 2015; Meleki,
Asiz, Smith, Gagnon, & Mohammad, 2011; Schneider et al., 2015; X. Zhang, Fairhurst, & Tannert,
2015) and methods to accurately model hybrid structural systems.
2.1.1 Reinforced concrete
Concrete is strong in compression and weak in tension; therefore, concrete is typically reinforced
for tension with steel (rebar). Flexural capacity of a section is developed through the composite
action developed between the steel reinforcing and the concrete. Reinforced concrete is as
composite material with typical compressive strengths from 25 MPa - 60 MPa; tension capacity
of reinforcing is typically 400 MPa - 500 MPa. Reinforced concrete can also have prestressed
tendons which can be much higher strength keeping significant tensile strains in steel tendons and
out of concrete. Cracked concrete occurs when the cracking modulus is reached and at that point
no concrete tensile capacity contributes to the strength and all tension is then carried by the steel.
7
Cracked concrete is significantly less stiff than uncracked concrete. This idea of cracked effective
stiffness is expanded on in the seismic design considerations section.
Reinforced concrete can have a variety of failure modes varying from very brittle to very ductile,
depending largely on the detailing of the reinforcing. As there has been much testing of reinforced
concrete for seismic design, ductile details are fairly well understood and routinely implemented.
One area of reinforced concrete detailing that is still an emerging technology is anchorage. The
CSA A23.3-14 still has this section as an annex, not yet mandatory; however, the section does
provide guidance and calculations to ensure ductility can be achieved in anchorage zones. Ductile
joints, walls, beams, and columns have all been thoroughly tested and proven and they are used as
required depending on the structural member being designed. Code prescribed ductility factors
(Rd) range from 1.0 to 5.0.
Seismic design of reinforced concrete is fairly well understood (Park & Paulay, 1975). Numerical
models can accurately predict global response of reinforced concrete structures (Lu et al., 2015).
In capacity designed concrete structures, ductility is achieved by allowing reinforcing steel to yield
in preferred locations, creating plastic hinges that will limit degradation of structural stability (Park
& Paulay, 1975). Also, concrete can be confined with reinforcement to improve ductility when it
is subject to crushing (SeismoSoft, 2014). Reinforced concrete non-linear behavior can be
categorized as either shear or flexural non-linear response.
2.1.2 Timber
Timber is anisotropic and hygroscopic meaning the properties vary with grain orientation and
moisture, respectively. Timber, as a material, can best be defined along three axes; longitudinal,
radial and tangential; however, timber design codes (CSA - CSA Standards, 2014b) do not
differentiate between the similar tangential and radial axis.
Timber can see 30 percent expansion radially and/or tangentially to a unit of longitudinal strain;
however, transverse strain in the longitudinal-radial or longitudinal-tangential plane results in
almost negligible (0.03) longitudinal strain (Green et al., 1999). For typical analysis, Poisson’s
ratio is taken as 0.3 because primary bending couples and/or axial compressive stresses will
generate this type of strain. It becomes increasingly important to consider Poisson’s ratio
8
variability in problems such as timber joinery where intimate wood contact is relied upon for
resistance or where unaccounted for wood strains may result in increased stresses that could lead
to brittle failures.
Shear modulus of timber in the longitudinal-radial plane is quite similar to that in the longitudinal-
tangential plane and is about a sixth of the longitudinal elastic modulus. The shear modulus in the
radial-tangential plane is much weaker and is only about 1/50-1/175 of the elastic modulus in the
longitudinal direction. While the stiffness is much lower in this direction, this plane does not
typically present a brittle failure mode. Shear in the radial-tangential plane would intersect the
wood fibres and result in wood crushing whereas shear in the other two planes would result in slip
between adjacent wood fibres which typically results in a brittle failure mode. Another potential
failure mode is rolling shear where wood fibres roll relative to one another under transverse shear.
Raw harvested timber has inherent variabilities. If a high grade of timber is desired, glue laminated
or other engineered wood products can be used to allow poor material to be removed and more
consistent quality materials to remain.
A distinction between light wood framing compared with heavy and mass timber is made. Light
wood framing covers the typical plywood or oriented strand board (OSB) and dimensional lumber
framing systems including light wood frame trusses and I-joists, typical for residential and other
low to mid-rise structures. Heavy timber typically refers to post and beam systems that are usually
either sawn timber or glue laminated or structural composite lumber. Mass timber typically
includes panelized products such as nail laminated timber, cross laminated timber or other
panelized structural composites. This study of hybrid concrete timber primarily focusses on mass
and heavy timber systems coupled with concrete as primary lateral system components; however,
one study includes a plywood diaphragm which falls into the light wood framing category.
Timber failure modes can be brittle under tension parallel to grain, tension perpendicular to grain,
and shear parallel to grain (CSA - CSA Standards, 2014b); therefore, timber connections, rather
than the timber itself, are typically desired to be primary sources of ductility (Braconi et al., 2007;
Fragiacomo, 2003; Ibarra et al., 2005; Schneider et al., 2015; Shen et al., 2013).
9
A high strength to weight ratio (FP Innovations, 2012) and increased damping characteristics
(Booth & Key, 2006) are advantageous for timber in seismic design. Ongoing research investigates
different types of wood materials including sawn (Chang et al., 2009), glue laminated (Popovski
et al., 2014), cross laminated timber (Yasumura, et al., 2015), nailed shear walls (Li et al., 2014),
and light wood framed diaphragms (He et al., 2011; C. Zhang, 2015). Code prescribed ductility
factors (Rd) range from 1.0 to 3.0 for wood framed systems.
2.2 Structural components
Specific structural components of a building work together to form a primary structural system
which carries the gravity and lateral loads. Lateral loads, particularly seismic induced lateral loads
are the focus of this study. Under seismic loads, specific components and areas of a structure are
designed to yield. Different seismic force resisting systems have unique costs and benefits. For
example, greater ductility reduction values typically require more effort during construction. Also,
components not part of the seismic force resisting system still require consideration with regards
to seismic design; therefore, the desired structural components, and the desired seismic force
resisting system structural components should be carefully selected.
2.2.1 Foundations and anchorage
Foundations and anchorage must be adequately designed and detailed, typically with reinforced
concrete, to transfer earthquake forces safely to the soil from the structure. For simplicity of
potential repair requirements, plastic hinges are often left to only form in the wall system,
maintaining an elastic foundation (Boivin & Paultre, 2012); however, foundation components,
such as a core wall extending into a basement, would still be subject to the ductile detailing
requirements of the SFRS above. Force controlled foundations or foundations designed for the
overstrength of their supported SFRS with capacity design principles may end up nearly twice as
thick as those designed only for seismic design force requirements (Klemenic et al., 2007).
A special study and other specific measures (NRC - National Research Council Canada, 2010)
may be required to better understand the interaction between the supporting soils and the
foundation and to satisfy the design criteria as outlined in Part 4 of NBCC (2015). Non-linear
foundation to soil interaction may need to be explicitly considered. Foundation-soil non-linearities
10
may include influences from both soils to footing and foundation to core. These non-linear
influences may be modelled with a foundation system which includes a rotational spring under
core elements, and a rotational and axial spring under basement walls (CSCE, 2008).
Foundations and soil interaction can play an important role in the performance of structures,
including consideration of whether to allow a foundation to behave more as a pinned base or fixed.
A footing which acts more as if were on pinned base supports has historically been referred to as
a rocking footing (Ventura, 2005). A rocking footing may reduce the force demand on a structure
but also increase drift. No rocking in the foundation is typically considered because the soil is
stiffer than the structure (Klemenic et al., 2007) and because many typical foundations will be
constrained in some way (Canadian Commission on Building and Fire Codes National Research
Council of Canada, 2015). Rather than rocking, foundations designed to CSA A23.3-14 are
currently classified as either capacity protected (fixed) or not capacity protected (rocking).
Foundations must be designed to remain elastic based on the capacity of the seismic force resisting
system (SFRS) unless specific detailing is provided to permit ductile rocking (Canadian
Commission on Building and Fire Codes National Research Council of Canada, 2015). Soil
interaction is typically covered in design by implementing site specific criteria that can be
determined based on geotechnical investigation.
In this study, implementation of site criteria rather than soil interaction will be carried out; no
special consideration of soil interaction is given beyond creating a site-specific response spectrum.
It is assumed that the foundation will provide a rigid base for the structure. The foundation and
anchorage design have been left for future study. Foundation displacements would need to be
added to the design drifts to determine if the superstructure design, coupled with the foundation
design, are to meet the performance criteria stated in the building code.
2.2.2 Timber frames
Timber sections are difficult to fix at the ends compared to leaving them with pin-end boundary
restraints except where they are continuous through multiple storeys. If fixing at ends is required,
timber will typically require some embedment into concrete or substantial steel fasteners. Timber
composite beams have been achieved with studs, screws, steel mesh or groove type connections
11
(Zhang et al., 2015). Timber in tension is a brittle failure mode and care must be taken to ensure
the capacity is not exceeded. Timber beams should be detailed to provide energy dissipation at the
connections.
In analysis of structures, if timber frame semi-rigidity increases the overall system stiffness by
more than 15 percent they must be included in the determination of the fundamental period;
whereas, if they are less than 15 percent, their contribution may be omitted. Two models must be
used to compare and determine whether semi-rigid timber frames should be included for period
calculations and to see if their performance helps reduce overall system requirements. Note that
even if they are not part of the SFRS, they still must be able to undergo system deformations that
are expected to occur during and after a design seismic event. Overstrength (Ro) and ductility (Rd)
factors for moderately ductile timber frames designed and detailed in accordance with CSA O86-
14 are Rd = 2.0 and Ro = 1.5. The intent of inclusion of semi-rigidity of timber frames in this study
is to see the impact of these frames and determine the most appropriate way to consider their
contribution (if any) to the SFRS.
2.2.3 Timber floor diaphragms
Diaphragms can be classified as rigid, flexible, or semi-rigid. Their classification impacts analysis
procedures, design requirements, and performance expectations. A rigid diaphragm may be an
appropriate assumption for a CLT floor system which is fairly stiff and may further have a concrete
topping; however, depending on the stiffness of the supporting shearwalls and connection
detailing, this may not always hold true. A flexible diaphragm would typically be considered if a
light wood framed diaphragm assembly is used without concrete topping (CSA, 2014), again
depending on the stiffness of supporting shearwalls. Detailed consideration of flexible plywood
diaphragms is provided in (Kleven & Noras, 2011). A semi-rigid diaphragm may be appropriate
classification for a CLT floor diaphragm; however, the computational efficiency is significantly
reduced when semi-rigid diaphragms are applied to a structural model (Computers and Structures
Inc., 2016).
Plywood and oriented strand board (OSB) diaphragms provide some rigidity. Envelope approaches
are recommended when stiffness cannot be accurately determined due to finishing such as concrete
12
toppings or ceiling materials. CSA O86-14 provides deflection equations for blocked shearwalls
(cantilevered vertical diaphragm) and simply supported horizontal diaphragms. These mechanics-
based approaches include contributions from chord elongations and slip, hold down slip, bending,
nail slip and panel shear (Association of professional engineers and geoscientists of British
Columbia (APEGBC), 2015). It is not typical to include the complexities of nailing and individual
panels into finite element models; however, Woodworks Shearwalls software for example will
capture this level of detail in structural models. Finite element models often use envelope
approaches, diagonal strut simplifications or other equivalent stiffness approaches. One approach
endorsed by the Canadian Wood Council is to follow ASCE 41-13 guidelines which define the
flexible versus rigid cut-off as when the maximum diaphragm deflection is more and less than
twice the average displacement of supporting vertical elements, respectively.
Diaphragm classification impacts design procedures for torsional response (NBCC, 2015).
Torsional response is increased with rigid compared to flexible diaphragms. As rigidity of a floor
system increases, force imparted on connections typically increases. Diaphragms can be designed
for inelastic performance but typically rigid diaphragms are not designed to yield. Classification
of concrete topped diaphragms may be affected by detailing of the topping. Adequate separation
should be provided to allow movement of the diaphragm or the topping must be designed to form
a part of the diaphragm. When concrete is added or the timber floor is of mass timber, the floor
system may be considered rigid (Isoda et al., 2016).
Mass timber floor products include nail laminated, cross laminated and structural composite
lumber panels. For nail laminated floor systems, a rigid diaphragm model coupled with a flexible
diaphragm approach to provide an envelope approach to lateral force distribution to vertical SFRS
elements is prudent but may be overly conservative so engineering judgement may still be required
to determine design forces (Binational Softwood Lumber Council, 2017). As the diaphragm forms
a part of the seismic force resisting system, internal diaphragm stresses must be calculated and
designed for as elastic components. Diaphragm to column connections must be checked to ensure
adequate overstrength is present to facilitate yielding of ductile components.
13
2.2.4 Shear walls
Shear walls can be constructed of reinforced concrete or timber elements. Where timber-based
shear walls are used, energy dissipation can occur within the shear wall boundary connections
(Schneider et al., 2015) (along with in the panel nailing in light wood framed shearwalls). For
timber shearwalls, boundary connection energy dissipation is desirable; however, current
connection materials would likely need to be repaired or replaced after a major event (Fragiacomo,
2011). Concrete shear walls may be better suited to taller structures where drift is a concern
because lateral displacement of a structure will be decreased when stiffer concrete shear wall
elements are included.
Concrete shear walls may have elements designed as piers, and spandrels. Analysis can vary
depending on if coupling, or partial coupling of piers is allowed and how involved of an analysis
method is used. The design methodology is analogous to slab edge-spandrel (L) or slab band (T)
beam design, where effective flanges of piers are considered in the design. Note the distinction in
terms that a spandrel in lateral core design is the lintel or coupling beam between two shearwalls;
whereas, a spandrel in gravity design typically refers to a slab edge (L) beam. Cantilever concrete
wall piers can be designed and detailed for seismic loading using ETABS (Computers and
Structures Inc., 2015) and design algorithms for CSA A23.3-14 are included in their shearwall
design manual (2015) to check the calculations. Wall ductility is directly related to the vertical
length of the plastic hinge, the horizontal length of the flexural compression zone, and the
compression strain capacity of the concrete (Adebar et al., 2005). Concrete core non-linearities
develop due to excessive strain in either material; the reinforcing steel or the concrete.
Appropriate modelling of concrete shearwalls may be achieved with link or fiber models. A
concrete shearwall fiber model may include multiple layers for reinforcing along with in-plane
shear and out of plane stiffness concrete layers, additional refinement of fibers near higher stressed
wall ends, out-of-plane bending stiffness reduction to 0.33EI, updated elastic shear properties with
assigned strengths, 25 percent increase for steel strain hardening, and 20 percent increase for
concrete compressive strength (f’c) due to confinement and other strengthening effects (CSCE,
2008). With a link model, automatic hinge properties can be defined according to ASCE 41
(Computers and Structures, 2015a).
14
2.2.5 Coupled shear walls
Coupled cantilever shear walls are cantilever walls connected by coupling beams. The coupling of
shearwalls induces opposing axial forces in coupled cantilever shear wall segments; the coupling
beam (also referred to as a spandrel or lintel) is the beam that induces these forces. The degree of
coupling ratio is established based on total system rotational resistance divided by the resistance
provided by uncoupled walls alone. The degree of coupling impacts the design ductility and
overstrength criteria. Coupled walls are stiffer than individual piers but not as stiff as solid walls.
Coupled walls are typical where door openings are required, such as in elevator cores. Diagonal
reinforcement is often used within the coupling beams to resist the large shear forces at beam ends.
Coupling beams with diagonal reinforcement have at times been found to be difficult to implement
in practice; as well, challenges with confinement of diagonal coupling beam reinforcing exist
(Klemenic et al., 2007). Quantifying elongation of diagonal coupling beam reinforcing for
transient coupling effects is important in some situations (Fox et al., 2014).
Design considerations include, beam to wall strength ratios, sequence of hinge formation, and
inelastic shear deformation of wall elements composing a wall system (Boivin & Paultre, 2012).
Consideration should also be given to displacement compatibility criteria to determine the
sequence of yielding and to provide a coupling beam stiffness of about 20 percent (Paulay, 2001).
One method to model coupling beam components includes definition of rigid end zones, shear
hinges (displacement type elements to model inelastic deformation of the coupling beam with
energy degrading hysteresis loops), and a vertical embedded beam element to transfer the coupling
beam moment to the wall (CSCE, 2008). Coupling beam compressive strains are often
underpredicted in NL-TH analysis so care in specifying ties for compression reinforcement is
required (Klemenic et al., 2007). Coupling beams may be modelled as frame elements, in which
case zero-length plastic hinge element equivalent to plastic hinge length for the beam-column
assembly may be used (Mousseau et al., 2008). Alternately, a designated length of plastic hinge
can be defined which helps to provide computational efficiency in some circumstances. Both zero
and finite plastic element length are preferred to a complex diagonally reinforced fiber model
approach (Wallace, 2012).
15
2.2.6 Non-components – components not part of the SFRS
Components not part of the seismic force resisting system (SFRS) either need to have sufficient
clearances to allow design drifts to occur without being influenced by the non-SFRS components
or the components must be able to serve their function during and after design drifts have occurred.
The timber connections validated in this thesis are an example of a connection that are able to
undergo significant drift and cyclic loading while maintaining some strength and stiffness. The
lack of separation of the timber beams to the SFRS means that the timber frames must be able to
maintain their function for gravity loads during and after predicted seismic induced displacements.
There is an option to include the timber components in the SFRS; however, the decreased R d and
Ro factors will result in higher design forces. The higher forces would be required because NBCC
requires the lowest Rd and Ro within the system, at location of interest, be used. This is where the
force-based design methodology of the NBCC does not account for unique system interaction and
provides a reason why studies of hybrid SFRS systems should be carried out. Other non-
components would include the remainder of the gravity system and structural components that are
not part of the primary structural system.
2.3 Seismic design considerations
In order to carry out appropriate analysis and design of the hybrid structure, specific seismic design
considerations were investigated
2.3.1 Ductility and overstrength
Ductility (𝜇) is the measure of absolute maximums of elasto-plastic displacement (𝛿𝑢 ) divided by
the yield deformation (𝛿𝑦 ) as shown in equation (2-1)
𝛿𝑢
𝜇= (2-1)
𝛿𝑦
Ductility is the primary reason for reduction of earthquake design forces. A structure that is ductile
will dissipate energy through non-linearity. This significantly reduces the force demand by
applying the equal displacement rule (for low to mid-rise structures) and reducing the force applied
but increasing the displacement incurred. Careful detailing is required to ensure ductility.
16
The experimentally verified timber frames of this thesis may be used as part of the SFRS if ductility
and overstrength factors (Rd, and Ro) are adjusted from those of the concrete core. The appropriate
Rd and Ro factors form a bit of a debate. The closest NBCC provided RdRo factors may be those
that are designed and detailed in accordance with CSA-O86 classified as moderately ductile
moment resisting frames with ductile connections. The timber frames are classed as much less
ductile system compared to the central concrete core wall system (Rd = 4 compared to Rd = 2). It
is not expected that including the stiffness provided by the perimeter timber frames will be
sufficiently useful to warrant inclusion as part of the SFRS due to the required increased design
seismic force on the structure as a whole; however, this will be explored. A fairly current resource
for determining RdRo is FEMA 695 (Applied Technology Council, 2009).
2.3.2 Structural irregularity
Structural irregularity is classified and separated into nine types in the 2015 national building code
of Canada. Each principal direction may exhibit a particular type of irregularity but not necessarily
will both directions have the same irregularities. Assumptions and unknowns regarding irregularity
are confirmed during the analysis and design procedures; however, to discuss by example how
irregularity impacts a structure, initial consideration for the structure of the extended study (as
shown in Figure 1-1) is provided as follows:
Type 1 – Vertical stiffness irregularity. The core walls (both coupled and uncoupled) have uniform
vertical geometric characteristics. This means that as long as wall depth reductions are not too
drastic moving towards the top, we should be able to meet the 70 percent of adjacent storey and
80% of adjacent three stories restrictions. Initial classification: Regular.
Type 2 – Weight (mass) irregularity. Our floors do not have weight irregularities and the slightly
lighter roof need not be considered. Initial classification: Regular.
Type 3 – Vertical geometric irregularity. The 130 percent benchmark will not be exceeded because
the horizontal dimensions of the shearwalls are equal moving through storeys and the storey
heights are equal. Initial classification: Regular.
17
Type 4 – In-plane discontinuity in vertical lateral-force-resisting element. This does not exist since
the SFRS is directly stacked. Initial classification: Regular.
Type 5 – Out-of-plane offsets - No out of plane offsets exist. Initial classification: Regular.
Type 6 – Discontinuity in capacity – weak storey. Since the capacity will only (if at all) decrease
moving up, there are no weak storeys. The foundation design must account for the strength of the
lowest storey or the foundation may become the weak storey. Initial classification: Regular.
Type 7 – Torsional sensitivity (to be considered when diaphragms are not flexible). With core wall
structures (such as the one in this study), torsional sensitivity often exists even if the center of mass
is reasonably well aligned with the center of rigidity. The torsional response occurs as a result of
eccentricities. Once further analysis is done, the parameter B will be calculated in accordance with
4.1.8.11.10) NBCC, and if it exceeds 1.7, then then Type 7 irregularity would exist. Initial
classification: Regular but pay attention for change in classification.
Type 8 – Non-orthogonal systems. Our SFRS systems in each principal direction are orthogonal;
therefore, Type 8 irregularity does not exist. Initial classification: Regular.
Type 9 – Gravity-induced lateral demand irregularity. This irregularity type is newly introduced
in the NBCC (2015). Essentially it reads as though inclined columns or floor cantilevers, if
significant, can cause the structure to shake more in one direction than the other. This may result
in increased drifts and likelihood of failure. Checks are further described in NBCC (A) - 4.8.10.5).
This structure does not appear to have this type of irregularity. Initial classification: Regular.
2.3.3 Torsion
Torsional aspects can be classified into two general categories; 1) natural (inherent) torsion which
is the eccentricity between center of mass and center of resistance or rigidity (plus dynamic
amplification), and 2) torsional moments due to accidental eccentricity. If the parameter B is less
than the limit of 1.7, equation 4-18 can be applied with the equivalent static force procedure;
however, if the limit is exceeded, a dynamic analysis is required.
18
The center of mass must be calculated. For a reasonably symmetric structure, the center of mass
is expected to be near the centroid in both x, and y coordinates. The center of rigidity must also be
calculated. This is done by calculating individual SFRS component stiffness’s and combining them
in such a way that all the deflections are compatible.
The approach to include the effects of accidental torsion depend on whether the structure is
torsionally sensitive or not. If B≥1.7 (torsionally sensitive): Two options exist in accordance with
4.1.8.12.(4)(a) NBCC 2015.
1) Shift the centres of mass of each diaphragm by ±(0.10)Dnx then superimpose the additional
forces due to torsion to the unrestrained 3D model forces. Then, take the maximum values
as elastic: storey shears, storey forces, member forces, and deflections.
2) Use the torsional moments calculated from the equivalent static force procedure but scale
them by (RdRo/IE) then superimpose them on the unrestrained 3D model. Then, take the
maximum values as elastic: storey shears, storey forces, member forces, and deflections.
If B≤1.7 (not torsionally sensitive): The approaches presented for torsionally sensitive structures
are still allowable; however, a shift by only ±(0.05)Dnx is required. Cases including those with
centres of mass un-shifted (only natural torsion exists), centres of mass (cm) shifted by ±(0.05)Dnx,
and cm shifted by ±(0.10)Dnx are analysed in this study.
2.3.4 Stability
P-delta effects are those effects that cause lateral demand on a structure due to gravity loads. P-
delta effects need to be accounted for in the design. Methodologies available are prescribed by the
2015 NBC. When gravity loads are also eccentric loads, P-delta effects increase significantly. P-
delta effects cause a negative slope on a force displacement plot which can lead to structural
instability if not effectively accounted for. Application of the stability coefficient (up to the
maximum limit of 0.4 prescribed by the NBCC (2015) is intended to ensure the force demand does
not exceed that of the resistance capacity, considering P-delta effects.
The relevant NBCC load combination for determining P-delta effects for seismic is given as
19
𝐶𝑎𝑠𝑒 5 = 1.0𝐷 + 1.0𝐸 + 0.5𝐿 + 0.25𝑆 + (1.0𝐶 𝑖𝑓 𝐶𝑟𝑎𝑛𝑒 𝑒𝑥𝑖𝑠𝑡𝑠) (2-2)
Following the recommendation of Filliatrault et al. (2013), the average displacements for the P-
delta check are to be considered (rather than the maximum); however, averages do include static
torsional effects.
Subclause 4.1.8.3.(8).(c) of NBCC (2015) requires sway effects due to interaction of gravity loads
with the displaced configuration of the structure to be included in the analysis. The NRC
Commentary (NRC, 2010) provides a stability coefficient that can be checked to determine if P-
delta effects can be ignored, shall be included, or if the structure is dynamically unstable. If the
stability coefficient at any level is less than 0.1, P-delta effects can be ignored at that level. If the
coefficient is greater than 0.4, the structure must be redesigned due to dynamic instability. The
stability coefficient (θx) equation, for each level, is provided as
𝑃𝑥 ∆𝑥
𝜃𝑥 = (2-3)
𝑅𝑜 𝑉𝑥 ℎ𝑠𝑥
The axial load, at the specific level, under the specified combination (Px) is multiplied by the code
specified drift limit (Δx) and is divided by the product of overstrength (Ro), shear force above and
including at that level (Vx), and the storey height (hs).
Notional loads (Nx) are those caused by initial material, section, and erection imperfections. The
steel design code requires that they be considered but concrete and timber codes currently do not.
As a hybrid structure, perhaps notional load inclusion would be prudent though it has not been
extensively investigated for this study. Note that the concrete design standard does specify a
minimum eccentricity for column loads (CSA - CSA Standards, 2014a); and the timber code also
requires that eccentricities be considered (CSA - CSA Standards, 2014b). A typical value of
eccentricity for a timber gravity loaded component would be 1/6 of the member width (Canadian
Wood Council and American Wood Council, 2016). This eccentricity of members influence on
global response is not investigated in this study. Notional loads (Nx) may be calculated as
𝑁𝑥 = 0.005(𝑃𝑥 − 𝑃𝑥−1 ) (2-4)
By including notional loads separately and P-delta effects through the stability coefficient,
structural stability is increasingly addressed.
20
2.3.5 Seismic weight
Weights of the various known superimposed loads and expected component self-weights are
tallied in accordance with Sentence 4.1.8.2 NBCC (2015); including, 60 percent of the storage
load, 25 percent of the snow load, and partition loads of 0.5 kPa (half the gravity design value).
The weights are typically lumped at floor levels. The lowest storey requires some additional
consideration because if the lower half of the lowest storey is lumped at the ground level, some
weight and resulting base shear is not captured. This is more important, yet still often neglected,
with heavy construction such as concrete or masonry (Drysdale, 2014); whereas, for a timber
structure the lowest half storey of shear may be more appropriate to neglect due to its relatively
low weight and increased flexibility.
2.3.6 Damping
Damping parameters need to be used with caution since there are many types of damping. Some
types of damping include hysteretic, acoustic coupling, energy radiation, and viscous, to name a
few (Irvine, 2016). Typically, for steel and concrete structures a damping value of five percent is
a good initial estimate (Booth & Key, 2006; Irvine, 2016) and for timber, it can often be found to
be much higher (Booth & Key, 2006). For a linear analysis, five percent damping is often what
structural design codes are based on (NRC - National Research Council Canada, 2010).
A suitable damping model could be the initial stiffness-based Rayleigh damping which is specified
at both first mode and mode number equal to N (Boivin & Paultre, 2012). There are varying
degrees of damping that can be experienced by concrete structures. Damping is reduced with
prestressing forces and increased with cracking of concrete (Booth & Key, 2006). A minimum
value of 2.5 percent is recommended for tall buildings which typically have lower values than low
to mid-rise buildings (Los Angeles Tall Buildings Structural Design Council, 2017).
For a non-linear response history analysis, damping may be calibrated experimentally from a
structure, such as the one in the three-dimensional study in this Thesis; otherwise, damping is often
set to two percent but should not exceed three percent of critical as per the NRC structural
commentary (2015).
21
Experimentally monitored values have been found to be between one and two percent (Boroschek
& Yanez, 2000). In this Thesis, for the shake table, three-dimensional experimental validation, the
hybrid concrete-timber structure with prestressing tendons, minimally cracked concrete, and
otherwise unfinished, damping was found to best fit near 2.2 percent (see Section 6.1.3, Three-
dimensional study, page 95). Based on the upper/lower bound limits described and the
experimental results noted, a reasonable definition of damping for a non-linear analysis of a mid-
rise concrete-core hybrid structure is between two and three percent.
2.3.7 Hysteresis
Hysteresis measures energy dissipation of a system by tracing load displacement or moment

rotation during cyclic loading. Typically, hysteresis is used at connection level but it can also be
used at a more global level. The area within the hysteresis trace equals the energy dissipated.
Energy dissipation is important in design of structures because energy dissipation reduces force
(load) demand on the system.
Hysteresis plots help analysts quickly identify hysteretic behaviour of the component; stiffness,
yield, plastic, and ultimate limits can be identified as well as degradation of stiffness through
repeated cycles. Compressive versus tensile behaviours are readily distinguished. Hysteretic
performance can be captured experimentally (Schneider et al., 2015) then numerically modelled
to represent actual connection behaviour. To numerically model hysteretic behaviour, researchers
use several different software packages such as SAP2000 (He et al., 2011; Meleki et al., 2011),
OpenSEES (Shen et al., 2013), SeismoStruct (Latour & Rizzano, 2015), and MARC.MSC (Xu et
al., 2009). When experimental results are well matched to a hysteretic model, the confidence of
the hysteretic model is improved.
Starting with the most basic and increasing in complexity, some prevalent hysteretic models are
reviewed. An elastic hysteresis model would simply follow a backbone curve over repeated cycles.
This type of hysteresis acceptable only for elastic design. A kinematic hysteresis model allows
loops to be formed. These loops indicate plastic deformation and energy dissipation. A degrading
hysteresis model provides reduction of strength and stiffness over subsequent load cycles. The
Takeda model allows for pinching. The Pivot model in and Pinching4 models provide similar
pinched, degrading trace capabilities (see Figure 2-1 (a) and (b), respectively).
22
(a) (b)
Figure 2-1 Hysteretic connection models (a) Pinching 4 [(Isoda & Tesfamariam, 2016) © with
permission from ASCE]; (b) Pivot hysteresis segment from ETABS
The combined effects of plastic deformation are captured by hysteretic models such as the Pivot
Hysteresis in ETABS and Pinching4 in OpenSEES. The Pinching4 model (Lowes et al., 2004) can
show good agreement for hybrid timber structure connections and timber connections (Schneider
et al., 2015) particularly when the ductility is expected to primarily occur in the timber and the
connector. Calibration values have been provided for some hybrid systems such as timber infill
frames using Pinching4 (Schneider et al., 2015; Shen et al., 2013).
Pinching in hysteresis is often recognized in timber systems. A cause for this behaviour is the
plastic deformation of wood. For example, in a nailed shear wall system, the nails will yield and
the wood will crush around the nails. The nails yielding is generally captured by a degrading
model; however, the wood crushing causes a loss of stiffness near the origin. This loss of stiffness
is due to the wood being fully crushed and not providing any load resistance within some region
under a subsequent load cycle. If timber is undergoing plastic deformation, the hysteretic model
chosen should include this pinching feature. Similarly, reinforced concrete exhibits pinching
behaviour; as steel reinforcement yields, concrete can crush locally and reinforcement can slip.
By using hysteretic connection models, a numerical structural model can capture the overall
picture of what is going on at specific connections without having to worry about capturing all the
details. A limitation of this approach is that different failure modes are not separately input and
assessed; proper connection design and detailing must be provided to ensure hysteretic behaviour
23
and subsequent failure modes are compatible. Pivot hysteresis is implemented with ETABS in this
study.
2.3.8 Capacity design
A strength hierarchy should be established for failure modes to ensure that the yielding elements
are able to provide the non-linearity required to provide adequate energy dissipation such that the
forces to be resisted by the structure can be reduced by corresponding ductility and overstrength
factors (Rd, Ro). The basic strength hierarchy or capacity design principles include prevention of
shear, anchorage or other brittle failures, maintaining stability of the system once plastic hinges
form, and protection of brittle regions of the system by ensuring their strength is greater than the
load that could be imposed by yielding members (Park & Paulay, 1975; Paulay & Preistley, 1992).
Material standards for both concrete and timber require that capacity design principles be followed
(CSA - CSA Standards, 2014a; CSA - CSA Standards, 2014b). By designing and detailing specific
elements and connections to behave in a ductile manner the designer can choose to maintain
strength and stability where it is most critical for overall structure performance while allowing
controlled yielding of specific structural elements and maintaining stability. Elastic (also called
capacity-protected or force-controlled) members should be designed with as little overstrength as
possible to ensure inelastic (also called plastic or ductile-yielding or deformation controlled)
members can be designed to be as light (economical) as possible. The elastic components are
designed to withstand probable forces that could be produced by inelastic (ductile) elements
(Filliatrault et al., 2013). A basic application of this is in a multi-storey frame structure where the
columns are designed as elastic elements and the beam is designed to yield. The yielding beam
allows for stability to remain whereas if the columns were to yield, global stability may be lost,
leading to collapse. This particular application of capacity-based design is known as the strong
column, weak beam principle, which is well used in current seismic design.
Another situation that may occur is that with existing structures which may be designed and
constructed prior to or not in accordance with modern seismic design codes. These structures will
often have less resilient structural systems. Structures that do not comply with current building
codes are more likely to have systems that take additional care and consideration to model. For
example, the weak points that may cause failure prior to intended locations of ductility will need
24
to expressly be either included or upgraded. Inadequate shear transfer elements, drag struts,
anchorage, or buckling failures may all lead to non-ductile failures that do not allow components
to behave in-elastically as intended. Whether modelling an existing or proposed structure, it is
important to ensure the likely failure modes are considered and the levels of ductility (or lack
thereof) and overstrength are properly accounted for.
For a tall shear wall structure, the walls are typically designed to experience yielding in the lower
(plastic hinge) portion. For a moment resisting frame, the beam will be designed to yield prior to
columns to help maintain stability. For somewhat of a combined system, coupled shear wall
assemblies use both flexural wall yielding (in the lower portion) and flexural coupling beam
yielding to provide ductility to the system. Various material models can be used to implement non-
linearity (Boivin & Paultre, 2012).
Alternately to determining forces on members that are designed to yield, the forces on non-yielding
(capacity protected) elements shall be determined based on the probable force imparted on them
by the yielding elements. For example, a building foundation will need to be designed to have
capacity in excess of that of the structure base to ensure the structure base is able to yield as
intended. In Canada, structural members will be either designated as force controlled or
deformation controlled then designed for either the forces due to loads or capacity of connecting
members and connections.
2.4 Code procedures and analysis methods
Non-linear behaviour of seismic force resisting systems (SFRS) plays an important role in modern
earthquake engineering. The 2015 national building code of Canada (NBCC) has specific
provisions and allowable methods for determining SFRS. Hybrid concrete-timber SFRS are not
presently included in the current building code.
Analysis methods considered here are those allowed according to NBCC (2015), which are
summarized in Table 2-1.
25
Table 2-1 Analysis methods prescribed in the 2015 national building code of Canada
Number Category Procedure Method
1 Linear Static Equivalent Static Force
2 Linear Dynamic Modal Response Spectrum
3 Linear Dynamic Response History
4 Non-linear Dynamic Response History
As well, response history analysis methods require the selection and scaling of ground motions.
This is an emerging science and for the first time the NBCC (2015) has included guidelines in the
commentary. This guideline as well as other current relevant material related to the selection and
calibration of ground motions is reviewed.
2.4.1 Equivalent static force procedure
The equivalent static force procedure (ESFP) is considered adequate for only limited structural
configurations. It must be validated as being an allowable option prior to completing design (NRC
- National Research Council Canada, 2010). The ESFP is a relatively simple static approximation
of a complex dynamic problem. ESFP is typically conservative when used within the code
specified limitations.
2.4.2 Modal response spectrum analysis
The elastic base shear is determined from the modal response spectrum method by combining
(through SRSS) maximum modal response. The elastic base shear (Ve) must then generally be
scaled to account for non-linearity (ductility) (Saaticioglu & Humar, 2003). This scaling will
impact the results by decreasing the base shear (V) and increasing drift. More specifically, when
using the NBCC (2015), further steps beyond just scaling for ductility are required to arrive at the
design base shear (Vd) (Filliatrault et al., 2013). Considering the equal displacement rule which is
generally relevant for short period structures (Filliatrault et al., 2013), scaling by overstrength and
ductility values help to provide more realistic anticipated drift and force values to be considered
in design (Canadian Commission on Building and Fire Codes National Research Council of
Canada, 2015). For forces, only one step of scaling is required. For deflections, the first step is the
same but an additional step of back scaling is required. The additional step is required because
26
deflections generally rebound and are therefore elastic. Residual (permanent) drift is not captured
in this form of analysis. Conversion from the elastic base shear (Ve) to the design elastic base shear
(Ved) is required. Experience has shown that short period, ductile structures have an inherent ability
to resist more load and further deformation, then that which is typically modelled (Canadian
Commission on Building and Fire Codes National Research Council of Canada, 2015); therefore,
the NBCC (2015) Clause 4.1.8.12.6 allows a reduction, provided the site class (soil conditions) is
not F, and that SFRS ductility is greater or equal to 1.5 (Rd≥1.5).
For strength, Clause 4.1.8.11.3. of NBCC (2015) places upper limits on the fundamental lateral
period (Ta) which are based on structural type. The limits are based on a series of tests and provide
a generally conservative envelope. Often the period determined by dynamic modal analysis will
be longer due to additional stiffening elements within the structure which are not part of the SFRS
and are not easily modelled. The code specified period (Ta,emp) is determined empirically.
2.4.3 Linear response history analysis
It has been noted that some researchers, engineers, and research groups use the term response
history and some use the term time history when discussing the same type of analysis. In this
thesis, the seemingly more current response history convention is adopted. No distinction is made
between the two. Response history analysis uses time-varying acceleration (ground motion) data
to determine the time-varying response of a structure. Response history analysis requires
complementary elements of structural and geotechnical engineering (CSCE, 2008). Geotechnical
aspects include developing the most appropriate ground motions while structural aspects include
predicting the influence of ground motions on the dynamic behavior of the structure.
A linear response history analysis carried out with ground motions tightly fit to the target response
spectrum should produce very similar results to a modal response spectrum analysis with the key
differences being that signs of forces are not lost. For example, combinations of axial and bending
loads and their contributions can be directly added rather than statistically combined (Charney,
2015). Goals of a response history analysis often include determining; 1) component inelastic
deformations, 2) design forces on force-controlled members, 3) storey transient and residual drifts,
4) floor acceleration and spectra, and 5) behavioral modes under significant non-linearity.
27
Component inelastic deformations, residual (permanent) drifts, and behavioral modes under
significant non-linearity are better predicted with a non-linear response history analysis. Multiple
response history analysis’ can be used to develop average, variability or likelihood of failures over
time. Goals can be evaluated either explicitly or indirectly (NEHRP, 2011). Intensity, scenario,
and time (risk) based are the three types of performance assessment (PA). Intensity-based PA has:
1) drift, and 2) force intensity. Scenario-based PA is used for: 1) average, and 2) variability. Risk-
based PA is a series of intensity-based results used to provide likelihood over time.
New guidelines for use with the 2015 NBCC for the selection and scaling of ground motions have
recently been released by NRC (2015). From these guidelines, Method A which is acceptable and
somewhat simpler than Method B1 and B2 may be used with response history analysis under the
NBCC (2015). Review of this ground motion selection and scaling procedures is provided after
the non-linear response history section to follow.
2.4.4 Non-linear response history analysis
Non-linear response history analysis allows for assessment of the local non-linearities and
calculation of time varying structural loads and deformations under cyclic input ground motions
(shaking). Response history analysis are typically used, along with peer review, to satisfy the
NBCC (2015) special study requirement which applies to systems not included in the list of code
defined seismic force resisting systems. This would apply to a hybrid concrete timber SFRS. Key
elements of a special study can be generalized to include compatibility of seismotectonic time
history, compatibility of inelastic properties, and interpretation of response.
Typical parameters of interest from a non-linear response history analysis include global response
parameters (shear, overturning and drift), and component responses such as moment/curvatures,
coupling beam rotations, permanent drift(s), changed periods of vibration due to non-linear
response, extent of damage and structural stability, ductility demand of local yielding elements,
and forces on non-yielding (elastic) components. Results from non-linear response history analysis
do not generally need to be scaled because they already should represent the non-linearities
associated with the structure (Filliatrault et al., 2013). The ability to more directly interpret results
from a non-linear response history is a key advantage to understand local performance, predict
28
order of failures, capture changes in fundamental period during non-linear response, and to better
predict stability under significant non-linearity.
Potential dynamic shear amplification caused by higher mode effects may not be fully captured in
results of a non-linear response history analysis. Additional allowance for increased shear demand
compared to that provided directly from interpreting the results of a non-linear response history
analysis can be critical to prevent un-anticipated brittle shear failures (Klemenic et al., 2007);
(Tremblay et al., 2001). This under-estimated shear demand is also common to modal and static
methods when predicting shear demand in tall shear wall structures (Panneton, Leger, & Tremblay,
2006).
The results obtained from a non-linear response history analysis should be reasonably similar to
those obtained from a properly conducted linear time history or linear response spectrum analysis.
Some software such as ETABS (CSI, 2015) have built-in post processing functions; however,
Matlab (Mathworks, 2017) or MS Excel (Microsoft, 2017) still may be useful for post processing
multiple analysis results (Boivin & Paultre, 2012). Post processing of results typically is required
or at least helpful to an analyst who must interpret non-linear response history analysis results. A
critical aspect of a non-linear response history analysis is peer review. A current consensus for
selection of peer review qualifications is provided by the Los Angeles Tall Building Structural
Design Council (Los Angeles Tall Buildings Structural Design Council, 2017).
2.4.5 Ground motion selection and calibration
Sourcing of ground motions, either historical or synthetic, is required for response history analysis.
Response history inputs must be pre-processed from their raw recorded data state then (typically)
calibrated to be representative of the site-specific design spectrum. Earthquake ground motions are
typically characterized by their frequency content, duration, intensity, and fault mechanism or
scenario. Ground motions for use with engineering analysis must be pre-processed including for
baseline correction and filtering and timestep (Bartlett, 2014; Boore, 2005; COSMOS, 2005;
Guorui & Toa, 2015). Orthogonal pairs of ground motions are required for bi-directional analysis.
Directionality and rotation to fault normal and fault parallel components is typically required when
less than 5km from a fault (Reyes & Kalkan, 2012); however, directionality of earthquake ground
29
motions may have a significant effect on the results (Archila & Ventura, 2012) and should be
considered with highly irregular or critical structures and those near to an active fault. Mainshocks
and aftershocks have unique implications on loss estimation due to earthquakes (Tesfamariam &
Goda, 2017) and aftershocks should not be used in isolation for a structural analysis but instead as
a follow up to the mainshock, if at all.
The site in this study is located near to the Cascadia subduction zone where three scenarios, crustal,
in-slab and interface, have been reported to contribute (Rogers et al., 2015). In-slab (within the
subducting plate and under the continental plate) and interface (at interface of subducting and
continental plate) events are part of the subduction process due to the Juan de Fuca plate subducting
under the North American plate; whereas, the crustal events considered are shallow and typically
within the North American plate. Data from Open File 8090 (Halchuk et al., 2016) is available
which deaggregates the seismic hazard by the scenarios described.
In 2018, the (2015) commentary to the NBCC was released which contains new guidelines for
selection and scaling of ground motions. Basically, Method A of the guideline requires that
scenario-specific ground motions be linearly scaled over scenario-specific period ranges. While
this new guideline is now released for Canada, a significant body of research is ongoing to develop
most appropriate methods for selection and scaling of ground motions for seismic design of
structures.
Historical ground motions contain real and ranging frequency content and represent real
earthquakes; they often provide good damage reference to previously damaged or destroyed
buildings; therefore, historical records are preferred when available (Naumoski et al., 2006).
Synthetic ground motions can provide a good spectral fit; however, they do not provide a historical
reference for damage and their spectral content is not necessarily compatible with scenario-
predicted events. The stochastic finite-fault method can be used to generate synthetic time
histories. For example, Atkinson (2009) developed several synthetic ground motion time histories
for uniform hazard spectrum (UHS) using this method. Lower period cut-off limits as specified
in NBCC (2015) were captured in this work. The steps to match simulated records to a target
uniform hazard spectrum are defined in Atkinson (2009).
30
Prior to even working on calibration of ground motions, it should be considered whether
calibration to fit the uniform hazard spectrum (UHS) is appropriate. Baker (2011) notes that the
uniform hazard spectrum can be overly conservative for ground motion calibration due to the
requirement to calibrate over the entire enveloped (worst case) acceleration response. Instead, the
conditional mean spectrum (CMS) may be more appropriate. This method scales back the UHS to
provide a more probabilistic scaling baseline for seed ground motions. Baker (2011) defines the
procedure to implement the CMS approach.
Calibration of ground motions refers to aligning aspects (e.g. PGA, S(T), etc.) to fit a specific site;
including both soil characteristics and nearness and directionality to various seismic hazards
(faults). It is agreed among researchers that ground motion characteristics used in analysis should
be calibrated to location or hazard specific conditions. It is not well agreed as how best to do the
calibration. Generally, there are two categories of calibration methods: 1) “scaling” (intensity-
based or weighted), or 2) “matching” (spectral).
Typically for ground motion calibration, matching refers to spectral matching; whereas scaling
refers to scaling some aspect of a ground motion to better match a component. A key limitation of
spectral matching is that you can only get a mean not a distribution (because the variability is
reduced during matching), this means that risk cannot be directly obtained (NEHRP, 2011, pp. 64).
ETABS Software (Computers and Structures Inc., 2015) can be used for matching by both
frequency and time domain; ETABS documentation (CSI Inc., ) provides scaling methodology
and discusses input options for historical time history files. Some specific scaling and matching
(calibration) techniques include:
1. Linear scaling (LS) – Uniformly scales the ground motion; types include:
a. PGA scaling – Scales the peak ground accelerations but neglects structure dynamics
b. Sa(T1) scaling – Scales to fundamental period but neglects higher mode response
c. S(Trsi) scaling – Scales to scenario-specific period ranges of interest
2. Sla Scaling – Scales to integrated response spectrum over a period range of interest
3. ASCE Scaling – As SIa except greater, rather than equal, integrated area is required
4. ATC Scaling – Procedure to eliminate variability between ground motions and account for
target spectrum and structural period (Applied Technology Council, 2009).
31
5. ATK Scaling – Initially selects the best naturally fitting ground motions then adjusts based
on target spectral acceleration divided by the response (Atkinson, 2005).
6. MSE Scaling – Minimizes mean square error between seed and target spectrum.
7. Modal-pushover based scaling – Considers structure strength (Kalkan & Chopra, 2010)
8. Frequency domain matching – Provides good spectral fit but poor dispersion of results;
also, alters non-stationary ground motion characteristics (CSI Inc., 2017; (Michaud &
Léger, 2014).
9. Time domain matching – Adds wavelets but keeps phase of seed ground motions
10. Hilbert-Huang transform – A spectral variability transform that preserves non-stationary
characteristics (Michaud & Léger, 2014; Ni et al., 2013; Barnhart, 2011)
One final type of ground motion scaling is scaling for importance. Importance of the system is
typically accounted for in a design model by increasing the seismic mass or by scaling the intensity
of the ground motion; this is a separate procedure entirely that happens later in the analysis
procedure.
32
3 Methodology
The results of the literature review indicate that hybrid concrete timber structures are increasingly
being sought and that a gap in meeting the demand exists. Further, for such structures, core wall
and coupled core walls with timber frame is a predominant structural form. Seismic analysis and
design are critical to developing new hybrid systems. Appropriate seismic analysis and design
requires consideration of individual structural components, and how to include the elements
selected. Factors such as ductility and overstrength, irregularity, torsion, stability, seismic weight,
damping, hysteresis and capacity design must be well implemented to develop a realistic numerical
model and resulting design. For such designs to go to construction, the relevant building code must
generally be followed. The national building code of Canada (2015) provides analysis methods
and guidelines which facilitate earthquake resistant design. Beginning with a linear static analysis
and moving to linear dynamic analysis methods then finally non-linear dynamic analysis improves
confidence in the results as the overall analysis progresses.
For hybrid concrete timber structures, there are no pre-defined and code specified hybrid seismic
force resisting systems. Some researchers are presenting suggested ductility and overstrength
factors for specific hybrid and timber systems; however, this is not the aim of this study. Instead,
this study aims to implement hybrid systems and look at the requirements to develop numerical
models that reasonably represent the structures as well as demonstrate analysis and design
procedures that are required and or recommended by the building code.
An appropriate way to demonstrate reasonableness (or correctness) in numerical modelling is

through examples where known quantities are validated. Software analysis reference manuals are
a good example of this methodology where numerous examples are verified to demonstrate
capabilities of the software. Similarly, experimental results from testing provide opportunity for
experimental validations. Many researchers use experimental data to fit hysteretic connection
models because connection hysteresis is a critical parameter in non-linear response of a structure.
A logical progression from simple connection validations to mid-rise structures includes

incrementally increasing the difficulty of the problem. For this reason, a two-dimensional study
was sought to ensure the model was behaving as expected when hysteretic connections were
included into a structural system. Next, a three-dimensional study was sought to establish
33
connectivity and dynamic properties. With all of these studies undertaken, the ability to develop
an extended study was realized. The extended study represents a possible structural system to meet
market demand for innovative hybrid concrete timber structures.
Based on the literature review, the aim of this research is to develop connection hysteresis for
timber and timber hybrid connections with an appropriate hysteretic connection model, develop
numerical modelling abilities adequate for two and three-dimensional structures, and finally to
extend the work to a useful type of in-demand mid-rise hybrid structural system.
3.1 Research planning
ETABS is a commercially available finite element software capable of implementing hysteretic

connection models and providing design information for structures. This software is used by many
researchers and for professional practice in this field; finally, an educational version was made
available; therefore, ETABS was selected to carry out the studies.
The experimental data was selected based on availability and meeting the research objectives. The
objectives being to incrementally move from connection calibration, to two-dimensional study, to
three-dimensional study, to application of the building code to an innovative mid-rise hybrid
concrete timber structure.
3.2 Examples chosen
Hysteretic connection performance was determined to be a critical parameter for non-linear

analysis of structures; the one-dimensional study provided opportunity to calibrate a hysteretic
model.
Interstorey drifts were determined to be a critical parameter determining structural response;

therefore, the two-storey model allowed for fitting of experimentally determined interstorey drifts
to a numerical model.
Damping and member stiffnesses were determined to have critical influence on structural
response; therefore, the three-dimensional study allowed for exploration of damping, diaphragm
34
stiffness and core wall stiffnesses while including hysteretic connection models for unique
connection types.
Following the building code was determined to be a critical requirement for structural analysis and
design; the extended study provided opportunity to apply the building code through analysis and
design iterations in developing an in-demand structural form.
3.3 Data collection
Data was collected through analysis of models developed in ETABS and post-processed in MS
Excel. Numerical models were saved uniquely at key steps in each process; this allowed a
collection of models to be kept for reference to critical phases of each procedure.
3.4 Importance and limitations
The importance of this research is that first connection models’ parameters are provided, second,
numerical modelling is demonstrated against experimental results, and third a structural design is
provided for an in-demand novel hybrid structure. Limitations are that the results are specific to
the exact arrangement studied and the work has not been parametrized to allow developing of more
broad conclusions
3.5 Proposed analysis
Based on similar research reviewed and discussed in the literature review, connection validations
will be carried out with a hysteretic connection model. Also, based on the literature review, two
and three-dimensional studies will be undertaken to implement seismic modelling approaches as
well as to ensure results are reasonably fitting the experimental data. Seismic analysis and design
steps were developed based on the seismic design flow chart provided in Figure 6.42 of Filliatrault
et al. (2013) including application of the NBCC (2015). Application of the NBCC (2015) is
appropriate because it reasonably represents the current state of the art, less the ongoing research.
In this study, basic minimum code procedures are followed but also the more extensive code
procedures that are optional but becoming more prevalent are followed which brings current
research in line with the code. Particularly with innovative structures, such as hybrid structures,
performing analysis beyond just the code prescribed minimums is recommended if not required.
35
3.6 Summary
The research methodology has been explained. The literature review identified key aspects to be
implemented including component types, seismic design considerations, and building code
analysis methods. The experimental validations provide a baseline for the extended study. The
extended study develops a seismic design of an innovative hybrid structure.
36
4 Case Details
Case details of the experimental validations and extended study are presented. These case details
form the basis of the remainder of work in this thesis.
4.1 Experimental validations
Case details for the one, two and three-dimensional experimental verifications are presented.
Experimental validations provide both a baseline for the numerical analysis (extended study) as
well as provide universal calibrated connection results for stand-alone use.
4.1.1 One-dimensional study
Reverse cyclic testing was carried out for timber connections by (Isoda & Tesfamariam, 2016).
The experimental apparatus with a representative timber beam is shown in Figure 4-1.
Figure 4-1 Experimental apparatus for connection testing [(Isoda & Tesfamariam, 2016) © with
permission from ASCE]
The input load protocol was defined as a function of yield rotation as shown in Figure 4-2.
37
Figure 4-2 Load protocol [(Isoda et al., 2016) © with permission from ASCE]
The tension bolt moment-resisting joint assembly includes a 38 mm x 38 mm steel square bar
embedded in timber with an end distance adequate to prevent pull-out. The three connections are
concrete core to timber beam, timber beam to timber column, and timber column to base. The
concrete core connection had a steel box at the core to receive the moment resisting joint. The
column base connection bolt went directly into a steel base plate and the beam column joint used
a steel bearing plate to prevent local crushing. The beam to core and beam to column connections
used (1) 16 mm bolt top and bottom; the timber base had (1) 16 mm bolt centred. The beam to
core and beam to column connections had three split rings while the column to base had just one.
All bolts were 450 MPa and timber was 10,500 MPa Young’s modulus, 30 MPa bending strength
glue laminated timber. Column sections were 170 mm square; beam sections were 120 mm wide
by 450 mm deep. General arrangements of the beam to core, column to base, and beam to column
connections are shown in Figure 4-3 (a), (b), and (c), respectively.
38
)
(a) (b) (c)

Figure 4-3 Connections (a) timber beam to concrete core, (b) timber column to steel base, (c)
timber beam to timber column [(Isoda et al., 2016) © with permission from ASCE]
4.1.2 Two-dimensional study
A two-dimensional semi-rigid parametric portal frame and its individual connections were
experimentally tested by (Mori et al., 2015). Their frame study was parametrized by beam depth.
The portal frame experimental test apparatus and reverse-cyclic load input is shown in Figure 4-4
(a) and (b), respectively.
(a) (b)
Figure 4-4 Portal frame study (a) portal frame experimental apparatus, (b) reverse-cyclic load
protocol [(Mori et al., 2015) © adapted for brevity with permission under creative commons
license (http://creativecommons.org/licenses/by/4.0/)]
39
Timbers were glue laminated Japanese cedar with Young’s modulus of 6500 MPa and bending
strength of 25.5 MPa. Column sections were 300 mm x 300 mm; beam sections were 240 mm
wide by 400 mm, 500 mm, and 600 mm deep. The portal frame had a 6000 mm beam span and
3140 mm column height.
Lagscrewbolt (lag-screw-bolt) fasteners were used. The lagscrewbolt fasteners consist of tubular
(hollow) lagscrews with coarse timber threads on the exterior and hex bolt threads on the interior.
Proprietary dovetailed steel plates were bolted to the lagscrewbolts. Beam to column and column
to base connection general arrangements are shown in
Figure 4-5. Static pushover analysis was used to verify global performance of the frames and
reverse cyclic loading was used to calibrate hysteretic connection performance.
(a)
(b)
Figure 4-5 Portal frame lagscrewbolt connections (a) beam to column connections, (b) column
to base connections [(Mori et al., 2015) © adapted for brevity with permission under creative
commons license (http://creativecommons.org/licenses/by/4.0/)]
40
4.1.3 Three-dimensional study
Full scale shake table testing was carried out on hybrid concrete timber structures (Isoda et al.,
2016). The testing parametrized the structure by number of storeys and diaphragm type. For the
three-dimensional study here, the two-storey concrete-core timber-frame hybrid structure from
(Isoda et al., 2016) was selected.
This two-storey structure utilized connections that were tested by (Isoda & Tesfamariam, 2016)
and are calibrated in the one-dimensional study in this Thesis. The general arrangement of the
structure is shown in Figure 4-6. Then, Figure 4-7 (a) and (b) show the shake table and hybrid
structure along with the load input to be considered, respectively.
Figure 4-6 General arrangement of concrete-timber hybrid [(Isoda et al., 2016) © with
permission from ASCE]
41
(a)
(b)
Figure 4-7 Three-dimensional two-storey concrete-timber structure (a) three-dimensional two-
storey concrete timber structure experimental setup (b) input load protocol [(Isoda et al., 2016) ©
with permission from ASCE]
The shake table loads were generated by the ground acceleration input and masses from gravity
loads of self-weight of the structural materials, an allowance for an 80 mm non-structural and
structurally isolated concrete topping on the plywood, and a uniform live load of 1.1 kPa.
The concrete core walls were 180 mm thick with 2000 mm openings on two sides. The pre-cast
concrete core was reinforced with 10 mm diameter, 400 MPa reinforcing spaced at 20 mm on
centre on each face in both horizontal and vertical directions. The core was fastened to the base
structure with four (4) and two (2) of 21 mm SBPR 930/1080 steel bars per wall with a prestress
force of 257.5 kN located as shown in Figure 4-8. The concrete core dimensions are provided in
Figure 4-8.
42
Figure 4-8 Concrete core section [(Isoda et al., 2016) © with permission from ASCE]
Timber beams, columns, and their connections were those defined already in the one-dimensional
study. Additionally, timber purlins were required to provide support for the diaphragm. The purlins
are estimated to be 120 mm by 300 mm placed at 1200 mm on centre. The diaphragm was 24 mm
thick JAS second grade structural plywood nailed with CN75mm nails at 50 mm on centre.
4.2 Extended study
The extended study structure is based on structural plan layout and storey heights provided in the
(MGB Architecture and Equilibrium Consulting, 2012) report Tall Wood. Elements of their 12-
storey FFTT (forest through the trees) system along with elements of their concrete base structure
are used to develop the preliminary structural system layout. Seismic design of buildings is
iterative and solutions can vary based on starting point and design decisions. The design solutions
of the extended study therefore do not match the design solutions of the original FFTT system but
rather use it as a starting point for analysis or stepping stone for the hybrid concrete timber primary
structural system of this study. Load conditions are also significantly different as the site of the
extended study is on soft soil in Victoria under high importance, whereas the original FFTT system
was for Vancouver on a rock site under normal importance. This difference results in roughly
double the seismic loading due to soil amplification, high importance factoring and higher initial
43
spectral accelerations. Further, the high importance criteria limit the interstorey drift to less than
half (2.5 percent is reduced to 1 percent) of that allowed under normal importance.
This section introduces the case details which were used to formulate the initial analysis model of
the extended study. Final case details are updated based on the results of the analysis and design
procedures.
4.2.1 Case details from 12-storey structural systems
Tall Wood (MGB Architecture and Equilibrium Consulting, 2012) presented a case for tall wood
structures. The structural aspects were authored by Equilibrium Consulting while the architectural
aspects were by MGB Architecture and Design. Equilibrium Consulting is a structural engineering
firm based in Vancouver, BC at the forefront of innovative timber design. One of the FFTT
structures in Tall Wood was the 12-storey timber structure. This structure also had a reference
concrete structure. Aspects of the concrete reference structure and aspects of the timber FFTT
structure were used to develop a starting point for analysis and design of the extended study 12-
storey concrete-core timber-frame (hybrid) structure. Nominally, the overall dimensions of the
structures allow for three 7.3 m bays in one direction and a 7.3 m, 9.9 m, and 7.3 m bay in the
other direction. Twelve storeys at 3 m per storey was typical for both the concrete base and the
Timber FFTT structures. Nominally 300 mm thick engineered timber slabs replaced 200 mm
concrete floors in the base compared to FFTT system. The FFTT system housed the hallways
surrounding the core within the core; whereas the concrete base building apparently did not require
as large of an overall core size. An additional column line was provided on each end of the core in
the concrete structure to carry gravity loads but the general layout of columns and beams was
otherwise similar. Steel beams were used in the FFTT system with reduced beam sections to
promote plastic hinging at preferred locations; primary non-linear response would however be
expected to occur within the core via panel connections and hold-down assemblies. Hinging in the
concrete base structure would presumably be limited to within the core coupling beams and lower
segment of core walls. Structural integrity reinforcing would be provided to prevent catastrophic
failures due to slab punching or other load effects. The general arrangement plans of the concrete
base and timber FFTT structures are shown in Figure 4-9, and Figure 4-10, respectively.
44
Figure 4-9 General arrangement plans (a) 12-storey concrete base structure [(MGB Architecture
and Equilibrium Consulting, 2012) © adapted for brevity and provided with permission under
creative commons license (https://creativecommons.org/licenses/by-sa/3.0/legalcode)]
45
Figure 4-10 General arrangement plan of 12-storey FFTT timber system [(MGB Architecture
and Equilibrium Consulting, 2012) © adapted for brevity and provided with permission under
creative commons license (https://creativecommons.org/licenses/by-sa/3.0/legalcode)]
The concrete base core walls were 350 mm in the cantilever direction and 450 mm in the coupled
direction, respectively. The cross laminated timber (CLT) core wall panels of the timber FFTT
structure were not provided for comparison. A concrete core elevation and plan were provided and
the core is shown in Figure 4-11. Loads considered for preliminary design in Tall Wood were
46
reasonably compatible with the loads required for the extended study; however, as discussed, the
loading in the extended study is pushed considerably higher by proximity to seismic source
hazards, soil conditions, and importance. The loading used for preliminary design in Tall Wood is
shown in Figure 4-12.
Figure 4-11 Core walls [(MGB Architecture and Equilibrium Consulting, 2012) © adapted for
brevity and provided with permission under creative commons license
(https://creativecommons.org/licenses/by-sa/3.0/legalcode)]
47
Figure 4-12 Preliminary load criteria [(MGB Architecture and Equilibrium Consulting, 2012) ©
adapted for brevity and provided with permission under creative commons license
(https://creativecommons.org/licenses/by-sa/3.0/legalcode)]
4.2.2 Case details for extended study initial analysis model
The initial analysis model for the extended study uses the same plan and storey dimensions as both
the FFTT and concrete base structures discussed in the previous section. The concrete core has the
same wall thicknesses as well. The timber connections are to utilize the connections from the one
and three-dimensional experimental verification studies. The change in vertical shear due to beam
spans and load change is assumed to have a negligible effect on the connection performance;
however, in a real scenario, this would need to be investigated and accounted for. Timber sizes
must be determined once the design process starts. A plan and isometric view of the initial analysis
structure was developed in ETABS and are shown in Figure 4-13 (a) and (b), respectively. Loads
for the preliminary design are provided in Table 4-1.
48
(a) (b)
Figure 4-13 Preliminary structural model (a) plan view with columns, beams, non-linear link
elements, floor planks, core walls and openings, (b) isometric view of the 12-storey hybrid
Table 4-1 Structural loads and deflection criteria
Climactic design data (NBCC 2015 and NRC Commentary, I-1, I-15)
Seismic Victoria, BC; High importance; Site class E.
Ground snow and rain load Ss=2.2 kPa,
Sr=0.1 kPa
Hourly wind q1/50=0.40 kPa (pressures vary)
Live and snow loads (NBCC 2015 and Table 4.1.5.3, NRC Commentary))
Roof Sroof=2.8 kPa Roof area: Is[Ss(CbCwCsCa)+Sr]
Floors 2-11 LFlr2-11=4.8 kPa (1.5[2.2(0.8)(1.0)(1.0)(1.0)+0.2)
Floor 1 (1st Storey) LFlr1=2.4 kPa
Dead and super-imposed dead loads (SIDL)
SIDL Roof Droof=1.5 kPa
SIDL Floors 2-11 Dflr2-11=2.5 kPa
SIDL Floor 1 (1st Storey) Dflr1=2.5 kPa
SIDL Cladding (all levels) Dclad=2.0 kN/m (perimeter with jogs at balconies)
Dead: All structural materials Self-weight (CLT floors, core, beams, columns)
Interstorey drift limit (NBCC, 2015)
Wind Hn/500
Seismic Hn/100 (with IE=1.5), reduced from Hn/40 typically
49
5 Procedures
This chapter outlines the procedures followed to complete the experimental verifications and
extended study.
Procedures to complete the one, two and three-dimensional studies are provided.
Fitted hysteretic models are required for non-linear analysis in accordance with the national
building code of Canada (2015). Strength, stiffness, degradation and pinching parameters
described in the literature review section are fitted using hysteretic plots. The procedure followed
for the one-dimensional study of hysteretic connection calibrations is explained. A typical link and
structural section object are shown in Figure 5-1; the link is tight to the left end (labelled as K1)
while the remainder is the structural section (labelled as C1).
Figure 5-1 One-dimensional model with hysteretic spring (K1) at global origin and structural
section (C1) with local axis extending from spring to joint restraint for rotational load input
The steps required to create the first analysis model are described then, briefly, the adjustments for
the subsequent connection models are discussed. In the models, the concrete core to timber beam
connection, loaded in the x and y-direction are noted as, BX and BY, respectively. For BX (strong
axis bending), the reverse cyclic load protocol did not reach six times the yield rotation prior to
failure; therefore, the models were calibrated to four times yield rotation only.
1. Create one-dimensional space with auto merge allowance for link elements
50
A typical cartesian (x,y,z) coordinate system was used and a grid in the z (global three) direction
was created. The final grid spacing equal to the height from the actuator to concrete core interface
(1330 mm) was achieved by first creating a 2 mm spacing then creating an adjacent 1328 mm
spacing.
2. Create section and zero-length link
The software’s auto merge tolerance was set to 3 mm to draw the link and elastic section. The
software’s auto-merge tolerance was reduced to 1mm forcing each end of the link to a single point
resulting in a zero-length lumped plasticity model.
3. Apply restraints to section for application of ground displacement load
The section was restrained at the free end, away from the link, in all but one degree of freedom;
namely, rotation in the direction of the applied load. This unrestrained rotation allowed rotation of
the timber relative to the connection interface, as would be experienced by the cantilever timber
(beam or column) in the test apparatus.
4. Create the load
A load pattern named displacement and a load case named non-linear direct integration history
were created. An arbitrary load of 1 mm translation was applied as a ground displacement joint
load to the top of the beam in the x (global one) direction. A time history function was defined
with the reverse cyclic load protocol presented in Figure 4-2.
5. Scale the load
Because the load input is in terms of rotation and ETABS does not allow a rotation load to be
specified, it was necessary to scale the load. For a small angle theta, the opposite side is a function
of the adjacent; therefore, the height from actuator to base of connection multiplied by the rotation
provided the displacement input required for the lateral movement of the restrained joint.
Therefore, a scale factor of (14.4 mm) was calculated as the height (1330 mm) multiplied by the
rotation at yield during the monotonic testing (0.0108 radians).
6. Define output time step and P-delta effects
51
P-delta effects were selected to contribute. The output time step was reduced to 0.01 seconds to
provide clearer hysteresis plots for visual verifications
7. Define the non-linear link element
The link element was defined with no mass or weight, as it is zero length. All translational degrees
of freedom were fixed because the connection was not allowed any gapping within the connection.
Axial rotation (R1) was fixed. Initially, the minor rotational degree of freedom (R2) was fixed;
however, when calibrating weak axis bending, this degree of freedom was modified. The major
rotational degree of freedom (R3) was selected to have linear and non-linear properties.
8. Define initial parameters of link elements
Monotonic load test data (Isoda et al., 2016) provided a yield moment and rotation of 24 kNm and
0.09 radians, respectively; with this, an effective stiffness of 2667 kNm/rad was calculated. The
value was decreased to 2600 kNm/rad after several fitting attempts were made and the connection
model was refined. The backbone curve was defined with the values provided in Table 6-1 and the
non-linear parameters were initially left as defaults.
9. Validate hysteretic model to experimental data through observation, trial and error
Post processing was done in MS Excel to allow overlay of test data and careful fitting of the results;
observations from the first connection calibration were as follows:
a) Increasing α tended to increase the horizontal spread of the hysteresis loops. α is defined
as a multiplier that locates the distance between the horizontal axis and the pivot point.
b) Increasing β tended to increase the pinching effect on the unloading of the hysteresis. β is
defined as a multiplier that pushes the unloading. While maintaining the envelope curves,
the energy dissipated in the connection is significantly decreased as pinching occurs. A
fully pinched model would have the β parameter equal to zero.
c) η was primarily ineffective in calibrating the connection model.
d) the unloading stiffness could not be modelled to be as stiff as the reverse cyclic testing
showed it to be. This is further discussed in the limitations section.
52
Final fitted parameters are shown in Table 6-1. In order to develop subsequent connection models,
adjustments were made to the Pivot model inputs, restraints, load orientation and magnitude and
section properties. Further observations made while fitting the remaining connection models were
as follows:
a) The pivot point was similar for both the strong and weak axis for this connection type;
possibly indicating the parameter alpha is constant for the connection type.
b) Pinching was less in the weak axis compared to the strong axis. The pinching is more of a
gap and the reduced effective depth of the moment couple (between the steel rods) tends
to require a smaller beta value.
c) The reloading stiffness of Pivot tends to degrade more than the experimental results.
d) The backbone curves match well.
e) The unloading stiffness heads through the axis using the Pivot point; whereas, the reverse
cyclic load tests unload almost directly over their load path. The resulting un-matched
unloading stiffness does not change the load envelope; however, the Pivot model would
predict increased energy dissipation due to larger area under the hysteresis loops.
10. Output hysteresis best fits
While fitting the hysteresis plots was a visual exercise and it is part of the procedure, it is also part
of the results; therefore, plots of best fit Pivot model hysteresis are shown in Figure 6-1 on page
91, for each of the four fitted connections, along with the tabulated results for connection parameter
inputs in Table 6-1, respectively.
A two-dimensional model was developed to ensure proper connectivity and modelling techniques
were being achieved in two-dimensional space and to provide Pivot model parameters for the
connections. Stiffness parameters for the backbone curves were developed from available reverse
cyclic test results (Mori et al., 2015). Timber sizes and properties were defined for the various
sections, connection types, and frame arrangements. Portal frame models were developed with
hysteretic connection (Ki) and beam (Bi) and column (Ci) sections as shown in Figure 5-2.
53
Figure 5-2 Two-dimensional non-linear portal frame model created with ETABS
First, individual components of the frame were calibrated according to experimental data (Mori et
al., 2015). This was completed using a non-linear static pushover analysis to develop the backbone
curves and with non-linear direct integration dynamic analysis to confirm the hysteresis. Because
ETABS does not allow joint restraints to be specific to a load case and to facilitate the two pushover
and response history analysis types, two separate models were created. Joint restraints were
assigned in the direction of the displacement load. Joint displacement was used for the dynamic
analysis because the load input is a time history acceleration. Conversely, the joint restraints were
not applied for a non-linear static pushover analysis; the load input is defined as a force on the
joint which was then scaled to achieve a predefined displacement. Initial calibration was estimated
from backbone curve data then fine tuning of non-linear stiffness was carried out visually using
MS Excel. Next, the static pushover curves for the frame were created. The frame height was
reduced to 2640 mm, a 500 mm reduction from the full height of the column defined in the
experimental model. This is acceptable because looking at the experimental setup, the top beam is
mounted at about a beam depth below the top of column and the columns were said to be 3140
mm tall. This adjustment to the work point in the model brought the stiffness of the frame to good
agreement.
Next, the non-linear dynamic analysis was considered. Here, the hysteretic properties of frame
connections were compared. The ground displacement load was applied to both top joints of the
portal frame with each of the top joints restrained in the direction of joint displacement. For final
load scaling, the rotation of the link at the base of the structure was matched to the rotation applied
to the connection calibration of the link element. An example of the hysteretic matching using
54
Pivot in ETABS is shown in Figure 5-3. With pushover analysis matching well, connection column
base hysteresis matching and no change in non-linear parameters from the previously fitted
connections, the goals of this study were achieved and so the remainder of the hysteretic fitting for
the portal frame was left for future study.
100
50
Moment (kNm)
-50
-100
-0.05 -0.025 0 0.025 0.05
Deformation angle (rad)
Experimental Pivot in ETABS HCB8-3
Figure 5-3 Hysteretic matching of column base in frame model
A three-dimensional study was undertaken to develop proper modelling and analysis techniques;
further, the three-dimensional study provided opportunity to implement fitted hysteretic
connection models from the one-dimensional study, providing further validation. The procedure
followed for experimental validation of the three-dimensional hybrid structure is presented in this
section. An isometric view of the model developed in ETABS is provided in Figure 5-4.
55
Figure 5-4 Three-dimensional model linear concrete core and non-linear timber connections
Concrete shear walls were modelled with linear shell elements and with pin connections to the
ground. Timber beams and columns were modelled with general linear sections. The columns are
continuous through the first level whereas the beams at the first level are not continuous.
Connections of timber elements to concrete and timber to timber, and timber to ground were
modelled with the Pivot multi-linear plastic element which were already calibrated in the one-
dimensional study. The plywood diaphragm was modelled with a membrane thickness of 24 mm.
The base acceleration was applied directly to all base nodes of the structure. The ground input file
length was 89.99 seconds with a time step of 0.01 seconds.
To test how close matching of the experimental versus finite element model was, two initial target
output variables were selected to match at the upper most nodes, furthest from the stiff concrete
core. Parameters to match were absolute maximum displacement, and peak floor acceleration; as
well, the dynamic response history of the structure was visually fitted considering phase and
frequency of motion.
In ETABS, there are many options for specifying damping properties including by individual
nodes. In this analysis, only the modal damping was adjusted. Based on the literature review, it
56
seemed reasonable to assume that the damping would be in the range of one to five percent for the
bare hybrid concrete timber structure. To help better fit the damping, a brief parametric study was
undertaken. Figure 5-5 shows the parametric study results and the matched floor displacement at
approximately 2.2 percent damping. This is a reasonable estimate of damping because; 1) the
structure is bare, 2) the bulk of the stiffness of the structure is due to the concrete core, 3) the
concrete core in the experimental test had prestressed coupling forces applied.
More than 100 iterations were carried out in fitting the model. Primary variables were the
diaphragm stiffness and damping of the structure. These variables would trade-off to achieve a
closer fit. For example, a good fit of peak displacement and floor acceleration could be achieved
with 0.6 percent damping and in-plane diaphragm modulus of elasticity of 6000 MPa; however,
the dynamic response was obviously underdamped compared with the experimental results. This
underdamping was apparent due to the excessive high frequency oscillations present in the
acceleration history output.
In another trial, the damping was increased and the diaphragm stiffness adjusted which resulted in
better matching dynamic response history and good matching peak displacement values but too
low of floor accelerations and not well fitted phase and frequency of displacements.
Effect of damping on maximum displacement

26
Maximum displacement (mm)
Matched displacement
22
18
14
10
1 2 3 4 5
Percent damping (%)
Experimental Finite Element Model
Figure 5-5 Parametric study of modal damping parameter
57
A secondary variable of concrete core cracking was introduced. Iterations with and without
cracked core were carried out. Implementing a cracked core did not improve the overall fit which
is consistent with the anticipated linear behaviour of the concrete core.
5.2 Extended study

Initial analysis and design steps the four (4) prescribed seismic analysis procedures of the national
building code of Canada (2015) are demonstrated for the 12-storey hybrid concrete-timber
structure considering material standards for concrete, timber, and steel (CSA - CSA Standards,
2014a; CSA - CSA Standards, 2014b; CSA - CSA Standards, 2014c). The general design
procedures and steps are based on Filliatrault et al. (2013). Also, just before the response history
analysis procedures, the procedure for selection and scaling of ground motions is demonstrated.
This procedure follows Method A from the newly released guidelines from 2015 NRC Structural
Commentary, Appendix to Commentary J.
5.2.1 Initial design steps
Prior to focussing in on any particular method of analysis and design, initial design steps are carried
out.
1. Determine basic seismic and site data
The building is to be located in Victoria, BC. Victoria was selected due to the relatively high
seismicity compared to many other regions in Canada (NRC - National Research Council Canada,
2010). The enacted building code for a site in Victoria, BC is presently the British Columbia
building code (2012) with Part 4 of NBCC (2010) adopted by the BCBC (2012). The BCBC has
not yet enacted a 2016 version; however, the 2015 NBCC is released and is being applied to this
Thesis. It is expected that soon, the BCBC will release a 2016 version that will replicate the
relevant section (4.1.8. Earthquake Load and Effects) of the 2015 NBCC. Two options are
available to determine the five percent damped spectral response acceleration (Sa) values for the
periods of interest at the site. They can be obtained from the NBCC (2015) Appendix C, or online
from the NRC Earthquake hazards website (Natural Resources Canada, 2015). In this study, the
parameters were obtained from the website because the exact latitude and longitude can be input
58
to provide an improved interpolation. The spectral accelerations for the code prescribed two
percent in 50 years probability are provided in Table 5-1.
Table 5-1 2015 National building code interpolated seismic hazard values
2%/50 years (0.000404 per annum) probability
Sa(0.05) Sa(0.1) Sa(0.2) Sa(0.3) Sa(0.5) Sa(1.0) Sa(2.0) Sa(5.0) Sa(10.0) PGA PGV
0.708 1.081 1.298 1.299 1.152 0.672 0.395 0.123 0.043 0.578 0.828
Assumed geotechnical parameters, averaged for the top 30 m of soil, include average standard
penetration resistance corrected to a rod energy efficiency of 60 percent of the theoretical
maximum (N60), average undrained shear strength (Su), and 3) average shear wave velocity (Vs)
as shown in Table 5-2.
Table 5-2 Geotechnical parameters provided for the site
Geotechnical parameters considered in this study

N60 10
Su 40 kPa
Vs 170 m/s
Site Class E
The bearing capacity and other soil parameters such as susceptibility to liquefaction and lateral
seismic soil loading on foundation walls shall be investigated during foundation design; however,
these are beyond the scope of this study.
2. Definition of importance (risk) category
Assumed architectural requirements indicate that the use of the structure is to be for emergency
services. This type of occupancy requires a seismic Importance (IE) of 1.5 because the structure
may be used in a post disaster situation. Note that while this classification requires an increase in
the primary structural systems ability to resist severe loading conditions, it does not necessarily
ensure operability and functionality for immediate occupancy (IO) which is the intent of the post
disaster importance category. If the intent of the post disaster importance Category is to be met,
other systems must adequately be designed to accommodate building movements under seismic
59
loading. The seismic importance factor (IE) not only increases the base shear (strength
requirements) but also tightens up interstorey drift (ISD, or drift) limits from 2.5 percent to one
percent.
3. Determine unusual factors and/or constraints
It was assumed that this post disaster building is to receive funding from the provincial government
which aims to increase the use of wood (timber) in public buildings (Government of British
Columbia, 2009). Further, (Association of professional engineers and geoscientists of British
Columbia (APEGBC), 2016) sustainability guideline requires that structural engineers consider
sustainability goals in their work. Increased use of wood has been proven to be sustainable as a
building material (Kremer & Symmons, 2015) so it ought to be considered for design; therefore,
the structure is to make use of wood components where suitable. It is anticipated that a hybrid
concrete timber structure may serve to meet the seismic demand requirement while simultaneously
meeting the intent of the Wood First Act (Gov. BC, 2009). An open layout is sought to facilitate
potential change in occupancy type and so a concrete core wall system with CLT floor panels and
perimeter timber frames is selected.
4. Check applicability of simplified approach and define the design spectrum
An update from the previous 2010 NBCC to the new 2015 NBCC is the new Section 4.1.8.1 which
allows a simplified approach to seismic design, given that certain criteria are met. This is the first
check in the seismic section of the building code. The check, to see if the simplified approach is
allowed, IEFaSa(0.2) and IEFsSa(2.0) must be less than 0.16 and 0.30, respectively. First though, the
site parameter (Fs) must be calculated; Geotechnical parameters must be used unless the structure
is to be founded on a rock site, as defined in NBCC 2015. Fs is determined with Equation 5-1.
1.0 𝑖𝑓 𝑟𝑜𝑐𝑘 𝑠𝑖𝑡𝑒 𝑜𝑓 𝑖𝑓 𝑁60 > 50, 𝑜𝑟 𝑆𝑢 > 100𝑘𝑃𝑎,

𝐹𝑆 = {1.6 𝑖𝑓 15 ≤ 𝑁60 ≤ 50, 𝑜𝑟 50𝑘𝑃𝑎 ≤ 𝑆𝑢 ≤ 100𝑘𝑃𝑎, 𝑎𝑛𝑑 (5-1)
2.8 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑜𝑡ℎ𝑒𝑟 𝑐𝑎𝑠𝑒𝑠
𝑁60 ≤ 15, 𝑎𝑛𝑑 𝑆𝑢 ≤ 50 ∴ 𝐹𝑠 = 2.8
60
𝐼𝐸 𝐹𝑠 𝑆𝑎 (0.2) ≤ 0.16𝑠, 𝑎𝑛𝑑
𝐼 𝐹 𝑆 (2.0) ≤ 0.30𝑠, 𝑖𝑓
𝐶ℎ𝑒𝑐𝑘, { 𝐸 𝑠 𝑎 (5-2)
𝑏𝑜𝑡ℎ 𝑡𝑟𝑢𝑒, 𝑢𝑠𝑒 4.1.8.1,
𝑒𝑙𝑠𝑒, 𝑢𝑠𝑒 4.1.8.2 − 4.1.8.22
(1.5)(2.8)(1.298) = 5.45 ≥ 0.16𝑠, 𝑎𝑛𝑑

(1.5)(2.8)(0.395) = 1.66 ≥ 0.30𝑠,
= {
𝑡ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒
𝑈𝑠𝑒 𝐴𝑟𝑡𝑖𝑐𝑙𝑒 4.1.8.2 − 4.1.8.22
Because the criteria for the simplified approach are not met, the code stipulates to move onto
Articles 4.1.8.2 to 4.1.8.22 of the NBCC (2015) to resolve the seismic force and force resisting
systems. The site coefficient (F) for spectral acceleration S(T) is determined in accordance with
Sentence 4.1.8.4.(4) NBCC (2015).
𝑆𝑎 (0.2)
(0.8)𝑃𝐺𝐴 𝑖𝑓 < 2,
𝑃𝐺𝐴
𝑃𝐺𝐴𝑟𝑒𝑓 = 𝑒𝑙𝑠𝑒 (5-3)
𝑃𝐺𝐴
{
1.298
= 2.14 > 2,
0.578
= 𝑡ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒
𝑃𝐺𝐴𝑟𝑒𝑓 = 0.578
{
Let Fa equal F(0.2), and let Fv equal F(1.0) in accordance with NBCC Sentence 4.1.8.4.(7). The
values from Table 4.1.8.4-B NBCC were used and are provided in Table 5-3.
Table 5-3 NBCC (2015) Table 4.1.8.4.-B

Table 4.1.8.4.-B Values of F(0.2) as a function of site class and PGAref
Values of F(0.2)
Site Class PGAref ≤ 0.1 PGAref = 0.2 PGAref = 0.3 PGAref = 0.4 PGAref ≥ 0.5
A 0.69 0.69 0.69 0.69 0.69
B 0.77 0.77 0.77 0.77 0.77
C 1.00 1.00 1.00 1.00 1.00
D 1.24 1.09 1.00 0.94 0.90
E 1.64 1.24 1.05 0.93 0.85
(1) (1) (1) (1) (1)
F
Notes to Table 4.1.8.4.-B: (1) See Sentence 4.1.8.4.(6).
61
Considering the PGAref value calculated earlier (0.578 g) and the Site Class (E), the value of F(0.2)
is taken as 0.85. Note that interpolation may be required but is not used here because 0.568>0.5
which is the largest table value. The remaining F(PGA, PGV, and T) values are not explained here.
Site specific spectral accelerations are calculated as shown in Equation 5-4.
𝑙𝑎𝑟𝑔𝑒𝑠𝑡 𝑜𝑓: 𝐹(0.2)𝑆𝑎 (0.2) 𝑜𝑟 𝐹(0.5)𝑆𝑎 (0.5), 𝑓𝑜𝑟 𝑇 ≤ 0.2𝑠
𝐹(0.5)𝑆𝑎 (0.5) , 𝑓𝑜𝑟 𝑇 = 0.5𝑠
𝐹(1.0)𝑆𝑎 (1.0) , 𝑓𝑜𝑟 𝑇 = 1.0𝑠
S(𝑇) = (5-4)
𝐹(2.0)𝑆𝑎 (2.0) , 𝑓𝑜𝑟 𝑇 = 2.0𝑠
𝐹(5.0)𝑆𝑎 (5.0) , 𝑓𝑜𝑟 𝑇 = 5.0𝑠
{ 𝐹(10.0)𝑆𝑎 (10.0) , 𝑓𝑜𝑟 𝑇 = 10.0𝑠
(0.85)(1.298) = 1.1 𝑜𝑟 (1.17)(1.152) = 1.348 ∴ 𝑆(0.2) = 1.348
(1.17)(1.152) = 1.348 ∴ 𝑆(0.5) = 1.348
(1.39)(0.672) = 0.934 ∴ 𝑆(1.0) = 0.934
𝑆(𝑇) =
(1.58)(0.395) = 0.624 ∴ 𝑆(2.0) = 0.624
(1.84)(0.123) = 0.226 ∴ 𝑆(5.0) = 0.226
{ (1.79)(0.043) = 0.077 ∴ 𝑆(10) = 0.077
F(PGA), and F(PGV) are selected as F(PGA)=0.74, and F(PGV)=1.17 from NBCC Tables. The
design spectral response acceleration S(T) is plotted in Figure 5-6.
Design response spectrum

1.6
1.2
S (g)
0.8
0.4
0
0 2 4 6 8 10
T (s)
Figure 5-6 Design spectral response accelerations
5. Determine the design control parameters (DCP’s)
Sometimes referred to as a seismic hazard index (SHI), design control parameters (DCP’s) are
code specified minimums used to trigger certain design criteria or system restrictions. The DCP’s,
and their limits for short and longer period accelerations (see Equation 4-5), are given in Sentence
4.1.8.6.(3), 4.1.8.10.(4) NBCC (2015), respectively.
62
𝐼𝐸 𝐹𝑎 𝑆𝑎 (0.2) ≤ 0.35 𝑡ℎ𝑒𝑛 𝑎𝑑𝑑𝑡 ′ 𝑙 𝑟𝑒𝑠𝑡𝑟𝑖𝑐𝑡𝑖𝑜𝑛𝑠 𝑎𝑝𝑝𝑙𝑦
𝐷𝐶𝑃′ 𝑠 = { (5-5)
𝐼𝐸 𝐹𝑣 𝑆𝑎 (1.0) ≥ 0.25 𝑎𝑛𝑑 𝑇𝑎 ≤ 1.0𝑠 𝑡ℎ𝑒𝑛 𝑎𝑑𝑑𝑡 ′ 𝑙 …
(1.5)(0.85)(1.298) = 1.66 ≥ 0.35 ∴ 𝑎𝑑𝑑𝑡 ′ 𝑙. 𝑟𝑒𝑠𝑡𝑟𝑖𝑐𝑡𝑖𝑜𝑛𝑠 𝑎𝑝𝑝𝑙𝑦

𝐷𝐶𝑃′𝑠 = {
(1.5)(1.39)(0.672) = 1.40 ≥ 0.25 𝑎𝑛𝑑 𝑇𝑎 𝑖𝑠 𝑢𝑛𝑘𝑛𝑜𝑤𝑛 ∴ 𝑤𝑎𝑖𝑡 𝑡𝑜 𝑠𝑒𝑒
Note that the DCP’s are not always governed by the same limits; as stepping through the code
provisions, one must specifically check the code to see which limits apply to which restrictions.
6. Load definition
With SFRS type defined Rd and Ro factors are initially assumed. Preliminary member sizing must
be completed to determine starting weights for the system. Loads required for the preliminary
gravity design are included in Table 5-4.
Table 5-4 Structural loads and deflection criteria
Climactic design data (NBCC 2015 and NRC Commentary, I-1, I-15)
Seismic See previous section
Ground snow and rain load Ss=2.2 kPa, Sr=0.1 kPa
Hourly Wind q1/50=0.40kPa (pressures vary)
Live and snow loads (NBCC 2015 and Table 4.1.5.3, NRC Commentary))
Roof Sroof=2.8 kPa Roof Area Is[Ss(CbCwCsCa)+Sr]
Floors 2-11 LFlr2-11=4.8 kPa (1.5[2.2(0.8)(1.0)(1.0)(1.0)+0.2)
Floor 1 (1st Storey) LFlr1=2.4 kPa
Dead and super-imposed dead loads (SIDL)
SIDL Roof Droof=1.5kPa
SIDL Floors Dflr=2.5kPa
SIDL Cladding (All Levels) Dclad=2.0kN/m (perimeter with jogs at balconies)
Dead: All Structural Materials Self-Weight (CLT Floors, Core, Beams, Columns)
Interstorey Drift Limit (NBCC, 2015)
Wind Hn/500
Seismic Hn/100 (with IE=1.5), reduced from Hn/40 typically
It is recognized that the loads shown in Table 5-4 are simplified, other specific point loads, guard
loads, etc. would need to be included during detailed design. For the design of the SFRS the loading
shown is adequate. Since the focus of this thesis is on seismic design not wind, wind load is omitted
but must be investigated for compliance with NBCC. Wind may govern either overall or in specific
63
storeys. Even if wind governs, seismic detailing must still be completed if one is to use the Rd and
Ro factors used to reduce the design base shear.
7. Determine basic gravity and SFRS components
Results of preliminary checks for the gravity system suggest that timber component sizes as shown
in Table 5-5 are adequate.
Table 5-5 Timber component sizes
Columns
C1 315 mm x 315 mm Glulam D.Fir-L 16c-E
C2 400 mm x 400 mm Glulam D.Fir-L 16c-E
Beams
B1 315 mm x 570 mm Glulam D.Fir-L 24f-E
B2 315 mm x 835 mm Glulam D.Fir-L 24f-E
CLT Floor and Roof
F1 315 mm thick E grade
Notes:
1. The larger columns may require glue-up in the shop.
2. The column sizes can reduce moving upward through the building
3. The beam sizes can reduce where live loads are 2.4 kPa
4. CLT floor plates consider Crosslam CLT Design Guide (Structurlam, 2016) and EQ (2012) report.
Since gravity design is not the topic of this thesis, the preliminary timber sizes shown will be used
for all levels. In a real design, it acknowledged that efficiencies should be realized.
8. Concrete core initial assumptions
Vertical SFRS components are of reinforced concrete. To account for decreased stiffness due to
cracked properties, the effective stiffness of concrete shear walls was initially assumed to be 0.35Ig
(as per the ASCE-7 requirement).
𝐼𝑐𝑟 = 0.35𝐼𝑔 (Initial Assumption to ASCE − 7) (5-6)
Further refinement considering axial load capacity demand ratio, in accordance with CSA A23.3-14
(Clause 10.14.1.2) was required and is carried out in subsequent iterations as follows:
𝐶𝑜𝑢𝑝𝑙𝑖𝑛𝑔 𝑏𝑒𝑎𝑚 𝑐𝑜𝑛𝑣𝑒𝑛𝑡𝑖𝑜𝑛𝑎𝑙𝑙𝑦 𝑟𝑒𝑖𝑛𝑓.: 𝐴𝑣𝑒 = 0.15𝐴𝑔 ; 𝐼𝑒 = 0.4𝐼𝑔 (5-7)
64
𝐶𝑜𝑢𝑝𝑙𝑖𝑛𝑔 𝑏𝑒𝑎𝑚 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙𝑙𝑦 𝑟𝑒𝑖𝑛𝑓𝑜𝑟𝑐𝑒𝑑: 𝐴𝑣𝑒 = 0.45𝐴𝑔 ; 𝐼𝑒 = 0.25𝐼𝑔 (5-8)
𝑊𝑎𝑙𝑙 = 𝐴𝑣𝑒 = 𝛼𝑤 𝐴𝑔 ; 𝐼𝑒 = 𝛼𝑤 𝐼𝑔 , 𝑤ℎ𝑒𝑟𝑒 𝛼𝑤 = 0.6 + 𝑃𝑠 /(𝑓 ′ 𝑐 𝐴𝑔 ) ≤ 1.0 (5-9)
Note that two distinct reinforced concrete shear wall systems are initially assumed to form the
primary vertical SFRS; ductile coupled walls and ductile shear walls. Ductility and overstrength
factors for ductile coupled concrete shearwalls and ductile shearwalls designed and detailed in
accordance with CSA A23.3-14 are Rd=4.0, and Ro=1.7, and Rd=3.5, and Ro=1.6, respectively.
Note that the system benefits from coupling action under loading in the x-direction only whereas,
in the y-direction, the core walls are not coupled but simply cantilevered. A check for degree of
coupling is required and takes the form
(𝑇)(𝑙)
(5-10)
𝑀1 + 𝑀2 + 𝑀3 + (𝑇)(𝑙)
Where (T) is the axial tension and compression acting at the centroid of the coupled walls which
is calculated as
∑𝑛𝑖=1 𝑉𝑓𝑏𝑖 =T (5-11)
Where Vfbi is the sum of the upward storey shears at the coupling spandrel edges of a given pier
and (l) is the distance between the centroids of coupled walls.
The degree of coupling must not be determined with a response spectrum analysis because the
reactions are not in equilibrium. Instead, the equivalent static force procedure (ESFP) was used to
complete and balance the check. It is presently assumed that full coupling action will arise from
the design. This assumption will be checked during ESFP analysis.
9. Determine the seismic weight (W)
The building loads which are used to develop the seismic weights were summarized in Table 5-4.
Note that the floor live load has been increased from 1.9 kPa for residential occupancy to 2.4 kPa
to allow for flexibility in occupancy allowing for office space for all floors above the first. The
first floor has been increased to 4.8 kPa to meet the lower storey office load requirement (NRC
Commentary, 2015). Assuming the weights of the perimeter timber frames fit within the allowable
1.0 kPa cladding load, the seismic weights are provided in Table 5-6. Later, the self-weight of the
timber frames is included.
65
Table 5-6 Preliminary seismic weights used for analysis
Wroof
Area (528 m2)[(1.0)(3.0)+(0.25)(2.8)]=1954 kN
Perimeter (92.8 m) (1/2(3)+1)(1.0)= 232 kN
Core (384 kN/m) (3/2)=576 kN
Sum Wroof=2762 kN
Wfloor
Area (528m2)[(1.0)(4.0)]=2112kN
Perimeter (92.8m)(3)(1.0)= 278kN
Core 384kN/m(3)=1152kN
Sum Wfloor=3542kN
The total seismic weight of the structure was estimated to be 2762 + 11(3542) = 41,724 kN.
10. Set the period (T)
Either, the empirical period (Temp) or a period (T) calculated according to mechanics in accordance
with NBCC (2015), within a range from empirical to a code specified upper limit (Tlim), can be
selected. There is a benefit to calculating the period (T), rather than simply using the empirical
period (Temp). If one assumes the analytical period is longer than the code specified period (Temp),
the preliminary base shear can be reduced. This is desirable because a lower starting point often
leads to a more economical design after some design iterations. The upper limits on analytical
periods (T), according to NBCC (2015) are provided to leave some allowance for stiffness
contribution of non-structural elements with the system (Canadian Commission on Building and
Fire Codes National Research Council of Canada, 2015). Where drift is governing, the period used
for these calculations can be a shorter (stiffer) period to possibly limit drift or alternately a longer
period to reduce force demand and also possibly reduce drift.
𝑇𝑒𝑚𝑝 , 𝑜𝑟 (5-12)
T= {
𝑇𝑒𝑚𝑝 ≥ 𝑇 ≥ 𝑇𝑙𝑖𝑚
3
{ 0.05(ℎ𝑛 )4 = 0.735 𝑠, 𝑜𝑟
0.735 s ≤ 𝑇 = 1.42 𝑠 (𝑢𝑛𝑐𝑜𝑢𝑝𝑙𝑒𝑑 𝑖𝑛 𝑥), 1.29 𝑠(𝑐𝑜𝑢𝑝𝑙𝑒𝑑 𝑖𝑛 𝑦) ≤ (2)𝑇𝑒𝑚𝑝 = 1.47 s
Since the period calculated with ETABS is less than the limit of double the empirical period and
less than an absolute value of four seconds, any value of period between the ETABS calculated
66
and empirical values can be used. It was anticipated that the perimeter moment frame may add
more than the 15 percent allowable stiffness so the period was initially taken as 1.0 s, and 1.2 s for
x, and y-directions respectively. The model for initial period determination was very limited,
including only cracked core walls, rigid diaphragms, and seismic weight applied as dead load with
the mass source switched to source the load patterns. Multipliers for the mass and weight of
materials were used to avoid counting their weights twice. The first and second mode period were
1.4, and 1.2 seconds.
The equivalent static force procedure (ESFP) allows for consideration of Temp, Tanalytical, or Tmax.
A longer period helps to reduce the demand on this initial design iteration. The period was set to
the maximum allowed based on the structure type and reduced drift limit of one percent for the
importance type.
11. Determine initial design base shear (V)
First irregularity types were considered; Type 7 may exist; Type 1 can be mitigated with thoughtful
design choices; and all other irregularities were not present in the structure. Note that according to
4.1.8.7.(1), if we cannot mitigate the torsional sensitivity (irregularity), the NBCC dynamic
analysis procedure must be used. Also note that Type 7 irregularity is not allowed for post disaster
structures, per 4.1.8.10.(2) NBCC. However, the ESFP provides a good initial starting point and
provides the base shear (V) that is required for subsequent phases of design.
5.2.2 Equivalent static force procedure
The equivalent static force procedure is implemented as follows:
1. Calculate base shear and storey forces, shears and overturning distributions
S(Ta) was interpolated for each direction (see Figure 5-6) and calculated as S(Ta,x=1.0 s)=0.934 g
and S(Ta,y=1.2 s)=0.872 g. Mv and J were calculated based on table 4.1.8.11 in the NBCC. First,
S(0.2)/S(5.0) ratio is found to be six. For the coupled system in the x-direction where Ta=1.0 s,
Mv=1.0 and J=0.97. For the wall system in the y-direction with Ta=1.2 s, Mv=1.01, J=0.93.
V= 𝑆(𝑇𝑎 )𝑀𝑣 𝐼𝐸 𝑊/(𝑅𝑑 𝑅𝑜 ) (5-13)
67
(0.87)(1.0)(1.5)𝑊
Vx = (4.0)(1.7)
= 0.192𝑊 = 0.19(41,724) = 8007kN
(0.93)(1.01)(1.5)𝑊
Vy = (3.5)(1.6)
= 0.24(41,724) = 10,014 kN
For walls and coupled walls, a lower limit is provided as

V= 𝑆(4.0)𝑀𝑣 𝐼𝐸 𝑊/(𝑅𝑑 𝑅𝑜 ) (5-14)
(0.3)(1.0)(1.5)𝑊
Vx = (4.0)(1.7)
= 0.066𝑊 = 0.066(41,724) = 2,754 kN
(0.3)(1.01)(1.5)𝑊
Vy = (3.5)(1.6)
= 0.08(41,724) = 3,338 𝑘𝑁
And since our system is ductile and soil seismic site class is E (better than F), an upper limit on
the seismic base shear is given as the greater of:
2
V= 𝑆(0.2)𝐼𝐸 𝑊/(𝑅𝑑 𝑅𝑜 ), and (5-15)
3
V= 𝑆(0.5)𝐼𝐸 𝑊/(𝑅𝑑 𝑅𝑜 )
2/3(1.348)=0.899, and S(0.5)=1.348; therefore the second equation governs and

Vmax=(1.348)(1.5)W/(RdRo)=0.297W, and 0.340W for x and y respectively. These upper limits do
not govern so the base shear of 8,007 kN and 10,014 kN were used for the initial ESFP procedure.
The base shear (V) is distributed over the height of the structure per 4.1.8.11.(7) NBCC. The
correction for taller structures (Ft) provided in the NBCC ESFP is calculated because the period is
greater than 0.7 seconds. Ft=0.07TaV giving 701 kN and 672 kN for x and y, respectively. Each
storey is assigned a fraction of the base shear according to
𝑛
𝐹𝑥 = (𝑉 − 𝐹𝑡 )𝑊𝑥 ℎ𝑥 /(∑ 𝑊𝑖 ℎ𝑖 ) (5-15)

𝑖=1
Overturning effects are determined with the following equation. J is modified to become Jx where
Jx=1.0 for hx≥0.6hn, and Jx=J+(1-J)(hx/0.6hn) for hx<0.6hn
𝑀𝑥 = 𝐽𝑥 ∑(𝐹𝑖 − ℎ𝑥 ) (5-16)
𝑖=1
Preliminary shear and overturning moment distributions, prior to inclusion of torsion and P-delta
effects, are presented in Figure 5-7 (a) and (b) show the system in the coupled direction resulting
in lower stiffness and lower loading whereas the y-direction (c) and (d) show increased storey
shear and overturning due to the stiffer cantilevered core walls.
68
Shear Overturning
12 12
8 8
Storey
Storey
4 4
0 0
0 5,000 10,000 0 100,000 200,000
Shear (kN) Overturning (kNm)
(a) (b)
Shear Overturning
12 12
8 8
4 4
0 0
0 5,000 10,000 0 100,000 200,000
Shear (kN) Overturning (kNm)
(c) (d)
Figure 5-7 Preliminary shear and overturning results (a) shear in x-direction, (b) overturning in
x-direction, (c) shear in y-direction, (d) overturning in y-direction
2. Consider accidental torsion, P-delta effects, notional loads and check drift limits
The eccentricity results and torsional characteristics of the structure as calculated with ETABS are
provided in Table 5.7.
Table 5-7 Eccentricities and total static torsional moment

Eccentricities (m)
ex Natural: |cmx -crx|=0.03 m, Accidental: 0.1(Dnx)=0.1(22.0)=2.20 m, Max total: 2.23 m
ey Natural: |cmy -cry|=0.00 m, Accidental: 0.1(Dnx)=0.1(24.4)=2.44 m, Max total: 2.44 m
Torsional Moments (kNm)
Tx T=Vex=(8007)(2.23)=17,856 kNm
Ty T=Vey=(10,014)(2.44 m)=24,434 kNm
69
Note that the torsional moments are split into a per floor moment and applied at each level. For
brevity here, only overall torsional moments are shown. Initially, the center of mass was displaced
in both principal directions as shown in Equation 4-18.
𝑇𝑥 = 𝐹𝑥 (±0.10𝐷𝑛𝑥 ) (5-17)
Next, the parameter Bx was calculated

𝛿𝑥𝑒,𝑚𝑎𝑥
𝐵𝑥 = 𝑤ℎ𝑒𝑟𝑒: 𝛿𝑥𝑒,𝑎𝑣𝑔 = 0.5(𝛿𝑥𝑒,𝑚𝑖𝑛 + 𝛿𝑥𝑒,𝑚𝑎𝑥 ) (5-18)
𝛿𝑥𝑒,𝑎𝑣𝑔
The model was first run with no eccentricities; then, the code specified eccentricity was applied.
The results for the x-direction were most severe in the first storey, where maximum displacement
was 2.4 mm and average was 1.3 mm, this results in
(2.4)
𝐵𝑥 = = 1.85>1.7
1.3
When loaded in the y-direction, the second storey is the most critical storey, By is given as
(2.6)
𝐵𝑦 = = 1.83 > 1.7
1.4
Based on these results, it is concluded that the preliminary structure is torsionally sensitive. Since
torsionally sensitive structures are not allowed for post disaster NBCC (2015), some modification
must be made. Maximum interstorey displacements were calculated as 12.3 mm and 13.7 mm in
x, and y directions were on the 9th and 8th floor respectively. Dividing by storey height (eg.
12.3/3000) gives drifts of 0.4 percent and 0.5 percent, then amplifying for non-linear effects (RdRo-
/IE), brings the drifts to 2.25 percent and 1.6 percent, respectively for x and y; the post disaster
limit of pne percent maximum drift was exceeded.
Stability coefficients were checked. Maximum stability coefficients occurred at ground level
where they were 0.01 which is considerably less than the limit of 0.1. This indicates that P-delta
effects need not be considered. While they need not be considered according to the code, it was of
interest; therefore, the effects were considered. P-delta effects were included as a non-linear load
case in ETABS such that all further analyses could then start from that point of P-delta modified
stiffness including self-weight. Since compressive stresses in axial-lateral systems increase sway,
the impact on the period provides a good point of reference. To illustrate the relative magnitude of
P-delta effects of the fundamental period (T), a non-linear case was defined where the period
calculation includes the P-delta effects. The results were that the period increased to 1.007 s from
70
1.000 s (less than one percent change). This small change in period helps further illustrates that
global sway effects are negligible.
Note that just because P-delta effects are negligible for global sway, P-delta effects on individual
members still need to be considered. Individual member design can typically use moment
magnification methods (CSA - CSA Standards, 2014a; CSA - CSA Standards, 2014b; CSA - CSA
Standards, 2014c) to account for lateral displacement of a member between its ends to account for
local (member-specific) P-delta effects. Because notional effects are currently only required for
steel, notional loads were not included in this analysis and are left for future study.
3. Wall coupling check
The wall base moments and the outermost couple are used to determine the percentage coupling
achieved. To achieve this, a two-dimensional model was used of only the coupling walls. It was
assumed that 0.3 of the ESFP loads were received for this check.
(10,123)(5.815)
= 0.79 𝑜𝑟 79% 𝐶𝑜𝑢𝑝𝑙𝑒𝑑 (5-19)
(2,515 + 10,383 + 2,520) + (10,123)(5.815)
Note that at least 66 percent is required to consider the concrete shear walls coupled (rather than
only partially coupled).
4. Update model properties
Model updates for cracked stiffness’s are calculated following A23.3 (2014),
𝐶𝑜𝑢𝑝𝑙𝑖𝑛𝑔 𝑏𝑒𝑎𝑚 𝑐𝑜𝑛𝑣𝑒𝑛𝑡𝑖𝑜𝑛𝑎𝑙𝑙𝑦 𝑟𝑒𝑖𝑛𝑓.: 𝐴𝑣𝑒 = 0.15𝐴𝑔 ; 𝐼𝑒 = 0.4𝐼𝑔 (5-20)
𝐶𝑜𝑢𝑝𝑙𝑖𝑛𝑔 𝑏𝑒𝑎𝑚 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙𝑙𝑦 𝑟𝑒𝑖𝑛𝑓𝑜𝑟𝑐𝑒𝑑: 𝐴𝑣𝑒 = 0.45𝐴𝑔 ; 𝐼𝑒 = 0.25𝐼𝑔 (5-21)
𝑊𝑎𝑙𝑙 = 𝐴𝑣𝑒 = 𝛼𝑤 𝐴𝑔 ; 𝐼𝑒 = 𝛼𝑤 𝐼𝑔 , (5-22)
607,000
𝑤ℎ𝑒𝑟𝑒 𝛼𝑤 = 0.6 + = 0.64 ≤ 1.0
30(350 ∗ 1000)
Note that the case shown was worst case (heaviest wall loading) and is at the base of the wall (per
A23.3); however, A23.3 allows averaging with less loaded walls. The parameter 𝛼 w was increased
to 0.6 for all concrete core walls (as an average).
71
Revision requirements are gathered and implemented to analysis models from findings of this
phase. A quick look at drift, torsional sensitivity, and period influence is provided in Table 5-8.
Table 5-8 Change in key parameters due to model update

Parameter Initial Updated
Bx 1.86 1.85
By 1.83 1.79
Rdx 4.0 3.5
Max ISD in x 2.25 3.1
MAX ISD in y 1.6 3.7
Tx 1.0 s 0.79 s
Ty 0.91 s 0.72 s
The fundamental period in each direction decreased by about 20 percent. It is noted that while
stiffness’ (EIeff) increased, so did the load attracted by the stiffer structure which significantly
increased predicted drifts.
5. Check if semi-rigid frames influence the SFRS
The semi-rigid timber frames were added to the model and checked for significant impact on the
structural performance, keeping in mind the 15 percent code limit on increased stiffness for
inclusion in the SFRS. It is assumed that construction detailing at the joints provides the same
timber sizes resulting in the same performance and that the change in vertical shear due to gravity
loads is negligible in the hysteretic performance of the connections. The addition of the semi-
rigidity to the timber frames provided only a 2.5 percent decrease in fundamental period. This was
lower than anticipated. It is noted that the 15 percent limit to additional stiffness is easily met; thus,
inclusion of the timber semi-rigidity in the SFRS is not required. It is not desired either since the
small additional stiffness will require approximately double the design force due to ductility R d
being reduced to maximum of Rd=2.0.
72
5.2.3 Modal response spectrum analysis
To achieve a passing design for global drift and shear in coupled beams, additional concrete core
wall thickness was required and two additional blade walls were also included in the core region.
The following steps were followed during the response spectrum analysis.
1. Perform response spectrum analysis
In each orthogonal direction and with the model restrained in the direction normal to the input, the
design response spectrum (S(T)) was provided as a load. From this analysis, the fundamental
lateral period (Ta), and the elastic base shear (Ve) for each of the two axes of consideration were
calculated as shown in Table 5-9.
Table 5-9 Modal response spectrum elastic base shear and fundamental period
x-direction y-direction
Ve (kN) 47,616 50,087
Ta (s) 0.767 0.697
2. Adjust to an elastic design base shear (Ved)
The elastic design base shear (Ved) was reduced by multiplying by two thirds of the response
spectrum at a 0.2 seconds period divided by the response spectrum at the fundamental period
(S(Ta)). Using the value of Ta calculated in the previous step, the following values of elastic design
base shear were calculated. Note S(Ta) is interpolated linearly based on the site-specific design
spectral response accelerations as shown in Figure 5-6. S(Ta,x)=1.127g, S(Ta,y)=1.185g.
2 𝑆(0.2) (5-23)
( ) ≤ 1.0
𝑆𝑖𝑡𝑒 ≠ 𝐹, 𝑎𝑛𝑑 3 𝑆(𝑇𝑎 )
𝐼𝑓 { ; 𝑡ℎ𝑒𝑛 𝑉𝑒𝑑 = (𝑉𝑒𝑑 ) ∗ 𝑔𝑟𝑒𝑎𝑡𝑒𝑟 𝑜𝑓
𝑅𝑑 ≥ 1.5, 𝑆(0.5)
≤ 1.0
{ 𝑆(𝑇𝑎 ) }
2 1.348
( ) = 0.797 ≤ 1.0
3 1.127
𝑉𝑒𝑑,𝑥 = (47,616) ∗ 𝑔𝑟𝑒𝑎𝑡𝑒𝑟 𝑜𝑓 { 1.348
}
= 1.196 ≤ 1.0
1.127
𝑉𝑒𝑑,𝑥 = 47,616𝑘𝑁; 𝑠𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦 𝑉𝑒𝑑,𝑦 = 50,087𝑘𝑁
Note that in the 2010 edition of the NBCC, the S(0.5)/S(Ta) term was not included; therefore, the
base shear would have been reduced in this step.
73
3. Determine the base shear (V); consider drift (Tdrift) versus strength (Tstrength)
For scaling, the NBCC (2015) equation used for calculating the minimum lateral earthquake force
(V) (Note: V=VESF) is used with unique period values for drift compared to strength calculations.
The base shear (V) has been calculated already in the initial design steps. For strength and drift,
the periods (Ta) are allowed to vary as Ta,drift and Ta,strength. In effort to not be overly conservative,
the period may be adjusted as follows for strength and drift calculations. Since T emp=0.735 s, and
Tx=0.767 s, Ty=0.697 s, use Tx=0.735 s, and Ty=0.697 s for both strength and drift calculations.
Therefore, S(Ta,x)=1.154, and S(Ta,y)=1.185. Finally, Vx=10,621 kN, and Vy=12,760 kN.
4. Calculate the scaling factor (Vd/Ve)
First, the design base shear (Vd) is calculated as shown in Equation 4-25.
𝐼𝐸 0.8𝑉 𝑖𝑓 𝑟𝑒𝑔𝑢𝑙𝑎𝑟 𝑜𝑟 𝑖𝑟𝑟𝑒𝑔𝑢𝑙𝑎𝑟 𝑏𝑢𝑡 𝐸𝑆𝐹𝑃 𝑎𝑙𝑙𝑜𝑤𝑒𝑑

𝑉𝑑 = 𝑉𝑒𝑑 ≤{ }
𝑅𝑑 𝑅𝑜 𝑉 𝑖𝑓 𝑑𝑦𝑛𝑎𝑚𝑖𝑐 𝑎𝑛𝑎𝑙𝑦𝑠𝑖𝑠 𝑖𝑠 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑
1.5
𝑉𝑑,𝑥 = 47,616 ≤ {10,621} = 12,004 𝑘𝑁 (5-24)
(3.5)(1.7)
1.5
𝑉𝑑,𝑦 = 50,087 ≤ {12,760} = 13,416 𝑘𝑁
(3.5)(1.6)
Then, the scaling factors (Vd/Ve)x,y are calculated as 12,004/47,616 = (Vd/Ve)x = 0.252, and
similarly (Vd/Ve)y = 0.268. These scale factors were applied to values obtained from the dynamic
analysis with unrestrained floors including: elastic storey shears, storey forces, member forces,
deflections. Deflections were also scaled once more.
5. Include effects of accidental torsion
Because the structure has B>1.7 in both x, and y-directions, and for simplicity, superposition of
the torsional moments due to ±(0.10)Dnx was used. Figure 5-8 demonstrates the effects of shifting
the centres of mass and the amount of inherent (natural) torsion. Note that there is negligible
inherent torsion. This is because the structure is very symmetric with centres of mass and rigidity
aligned. In this particular case, it is demonstrated not to be necessary to superimpose the 0.10Dnx
values to the restrained model because the restrained model is equal to the zero-eccentricity model.
74
Varying eccentricity and check inherent torsion
12
8
Storey
Note: Restrained model and zero

4
eccentricity plots overlap, indicating
negligable inherent torsion
0
0 200,000 400,000 600,000 800,000
Storey Torsion (kNm)
e=0 e=0.05 e=0.10 restrained to only alow x,z reactions
Figure 5-8 Varying eccentricities and inherent torsion
6. Convert from elastic to inelastic forces and pseudo-elastic displacements
Both elastic and scaled response spectrum x-direction storey shears are shown in Figure 5-9. The
figure illustrates the impact of including ductility into seismic design by showing the magnitude
of force reduction achieved. Figure 5-10 shows the scaling method for drift. Note that drift results
were exceeding the specified drift limit for post disaster; therefore, design iteration was required.
Scaling of elastic forces

12
8
Storey
0
0 2000 4000 6000 8000
Storey forces (kN)
Elastic Values Scaled by Vd/Ve
Figure 5-9 Scaling of elastic forces
75
Scaling of elastic drifts considering torsion Elastic Drift e=0
12 Elastic Drift e=0.05

Varying
eccentricities
1 Elastic Drift e=0.10
Initial
Elastic
Elastic Drift e=-0.05 (neg)
Values
8 2 Drift Scaled with (Vd/Ve), e=0
Scaled
back by 3 (final)
Storey
(Vd/Ve) Drift Scaled with (Vd/Ve), e=+0.05

Back scaled
by (RdRo/IE)
Drift Scaled with (Vd/Ve), e=0.10
4 Drift Scaled with (Vd/Ve), e=-0.05 (neg)

Post disaster drift limit
Drifts Back Scaled by RdRo, e=0
Drifts Back Scaled by RdRo, e=+0.05
0 Drifts Back Scaled by RdRo, e=+0.10

0 0.005 0.01 0.015 0.02
Drift Drifts Back Scaled by RdRo, e=-0.05 (neg)
Figure 5-10 Scaling of drifts considering torsion
5.2.4 Selection and scaling of ground motions
The selection and scaling of ground motions for response history analysis following Method A,
from the appendix of Commentary J in the NRC Structural Commentaries (2015), was applied for
the site and structure as follows:
1. Determine the target response spectrum S(T)
The response spectrum S(T) used for ground motion selection and scaling procedures was the site
modified, and low period detailed spectrum shown in Figure 5-11.
76
Site modified response spectrum
1.6
S (g) 1.2
0.8
0.4
Low period detailed
0
0 0.4 0.8 1.2 1.6 2
T (s)
NBC Modified S(T) Site C NBC S(T) for Site Class E
Modified S(T) Site Class C Spectrum
Figure 5-11 Response spectrum for selection and scaling of ground motions
2. Determine the period range of interest (TR)
The period range of interest was governed by the 90 percent participating mass ratio on the low
spectrum and the minimum upper limit for the high end of the spectrum. Participating mass ratios
from contributing modes are shown in Table 5-10 and the resulting period range of interest is
shown in Figure 5-12.
77
Table 5-10 Participating mass ratio's
Participating mass ratio (%)
Mode T (s) X Y Z Mode
1 0.655 0% 34% 2% X,1
2 0.608 28% 34% 13% Y,1
3 0.577 32% 35% 83% Z,1
4 0.193 33% 35% 92% Z,2
5 0.154 72% 35% 92% X,2
6 0.151 73% 67% 92% Y,2
7 0.113 73% 68% 96% X,3
8 0.080 73% 68% 97% Y,3
9 0.074 86% 68% 98% X,4
10 0.068 86% 80% 98% Y,4
11 0.061 86% 80% 98% Z,3
12 0.050 86% 80% 99% Z,4
13 0.049 92% 80% 99% X,5
14 0.043 92% 87% 99% Y,5
15 0.042 92% 88% 99% Z,5
16 0.037 95% 88% 99% X,6
17 0.036 96% 88% 100% X,7
18 0.032 96% 88% 100% Z.6
19 0.030 96% 92% 100% Y,6
Define TRS
1.6
1.2
S (g)
0.8
0.4
0
0.01 0.03 0.1 1 1.5 10
T (s)
Spectrum Spectrum within TRS
Figure 5-12 Period range of interest
78
3. Determine the seismo-tectonic regime
Because the site is located near the Cascadia subduction zone, three earthquake hazard scenarios
were selected for the target ground motion suites; namely, crustal, in-slab and interface
earthquakes. The nearest point from the earthquakes Canada hazard calculator (Natural Resources
Canada, 2015) was referenced to choose the Latitude and Longitude to be matched with Open File
8090. Grid points are valid within about 10 km which is less than the closest grid point at 3.9 km
away; therefore, there was no need to interpolate data file values. Data from Open File 8090
(Halchuk et al., 2016) was collated and plots were developed to demonstrate trends of the scenario-
specific hazard contributions as shown in Figure 5-13 (a) and (b) with deaggregated spectra and
relative contributions, respectively. Further, deaggregation plots requested from Earthquakes
Canada and received via email (Canadian Hazards Information Service, 2017) were inspected to
identify dominant scenarios. Selected deaggregation plots within the period range of interest and
at the probability used in the 2015 NBC are provided in Figure A-1, Figure A-2, and Figure A-3
of Appendix A.
Deaggregated spectra Relative contributions

1.50 0.80
Scenario relative hazard
0.60
1.00
S (g)
0.40
0.50
0.20
0.00 0.00
0.01 0.10 1.00 0.01 0.1 1 10
T (s) T (s)
Crustal In-Slab
Crustal In-Slab Interface
Interface UHS 2%/50yrs
(a) (b)
Figure 5-13 Deaggregated hazard (a) deaggregated and uniform spectra, and (b) relative
contributions of scenario-specific hazards
4. Define scenario-specific magnitude distance (M-R) relationships
The magnitude distance relationships for each scenario must be defined. These magnitude distance
relations, along with soil parameter Vs30 formed the primary basis of ground motion record
79
selection while other secondary factors such as spectral shape, duration, and size of scale factor
formed the secondary part of the ground motion selection process. To estimate the dominant
magnitude-distance scenario’s, global mode, global mean, and inspection of deaggregation plots
methods were completed and are provided as additional information in Table A-1 and Table A-2
of Appendix A. The final target scenarios based on deaggregation inspection are presented in Table
5-11.
Table 5-11 Target scenario's
Mw Rhyp (km) Vs30 (m/s)
Crustal 6.8 10
In-slab 7.4 70 100-250*
Interface 9.0 50
* Ideally Vs30 would be 0-180 m/s (site class E) but practically nothing below 100 m/s exists and some slightly
over 180m/s may be well worth keeping as the values are not absolute; a strict interpretation of site class would be
overly restrictive.
5. Scenario-specific period ranges of interest (TRS,i)
With guidance not provided in the 2015 NRC Commentary on how exactly to develop scenario-
specific period ranges, initially a 30 percent relative hazard assignment was attempted; however,
this led to excessive scale factors due to maintaining not less than 90 percent of the mean over a
longer segment of the period. Upon further literature review (Daneshvar, Bouaanani, Goda, &
Atknison, 2016; Dehghani & Tremblay, 2016; Gonzalez, 2017; Monroy, Hull, & Atukorala, 2016;
Shafiq, 2018; Tremblay et al., 2015) and review of deaggregated spectra dominant scenarios, the
final scenario-specific period ranges of interest were adjusted with judgement to those shown in
Table 5-12.
Table 5-12 Scenario-specific period range of interest

Scenario Period range of interest
Crustal 0.03 s ≤ TRS,crustal ≤ 0.40 s
In-slab 0.03 s ≤ TRS,in-slab ≤ 0.80 s
Interface 0.30 s ≤ TRS,interface ≤ 1.5 s
Exact methodology for defining the period range of interest was beyond the scope of the 2015
NRC Commentary but should be investigated further because the ground motion scale factors will
directly affect the structural response.
80
6. Selection of ground motions
The selection of ground motions required utilization of multiple databases and supporting
documentation on a scenario-specific basis. The PEER NGA-West2 database (Ancheta et al.,
2018) provided records compatible with the crustal hazard. Using selection criteria of Vs30 less
than 180m/s, distance (Rrup) less than 50 km, and magnitude (Mw) of 5.5-7.5, initial selection was
performed. To further reduce the number of ground motions, the design spectrum was considered
to allow minimizing the mean square error. Duplicate records from the same event were removed
and scale factors were minimized for scaling to Trs,i. The K/KiK-Net database also contained an
additional crustal record which made it into the final 11 selected records but was removed prior to
the final three suites of five bi-directional ground motions (see Table A-3, Appendix A).
The K/KiK-Net (NIED, 2018) and Universidad de Chile (Departmento de Ingenieria Civil, 2018)
databases were used to source subduction type events, including both in-slab and interface events.
Flat files from Dawood, Rodriguez-Marek, Bayless, Goulet, & Thompson (2016) and Bastias &
Montalva (2016) were used to assist with the selection processes for Japanese and Chilean
earthquakes, respectively. For in-slab events, initial priority was given to soil parameter (Vs30)
which resulted in the need to relax the magnitude and distance criteria. To increase the data pool,
stochastic time histories from Atkinson (2009) M7 accelerograms were also considered; however,
due to the preference of historical records and the criteria to minimize scale factors, these stochastic
time histories were not included the final in-slab suite.
Ground motions originating from Japan’s Kik-Net and Chile’s Universidad de Chile databases
were utilized for interface events. Retrieval of ground motion data was facilitated through
COSMOS VDC (CESMD, 2018) as well as source databases. Simulated ground motions from
Atkinson (2005) and Atkinson & Macias (2008) were also considered. The simulated ground
motions were considered high quality for meeting the target magnitude distance scenario but lower
quality compared to historical records in confidence of input time-acceleration values. None of the
stochastic records made it to the final interface suite. Final selection of ground motions did not
occur without going through the pre-processing and scaling procedures. The selection process was
somewhat iterative as the final selected ground motions were refined by reviewing the magnitude
of the required scale factors. Final selected ground motions are presented in Table 6-3 along with
scaled mean spectra results are provided in Figure 6-13, on page 101.
81
7. Pre-processing of ground motions
Pre-processing requirements were primarily dependent on the source database. PEER NGA-West2
provided pre-processed accelerograms and any additional filtering was found to only distort the
ground motions. Baseline correction was however at times required as integration of the
accelerograms sometimes resulted in non-zero residual displacement trends. K/Kik-Net provided
uncorrected (not pre-processed) accelerograms; therefore, filtering and baseline correction was
always required. Universidad de Chile ground motion files were provided in both corrected and
uncorrected formats. Where suitable, pre-processed ground motions were utilized. Where filtering
was required, a Butterworth bandpass filter was used with frequency cut-offs at 0.1 Hz and 25 Hz
for low and high frequency, respectively. When baseline corrections were required, a linear
correction was first attempted; if the linear correction was not providing acceleration, velocity and
displacement traces that best matched the original signature amplitude and phase, the filter order
was increased to quadratic then cubic if required. Accelerogram timesteps varied from 0.004 to
0.01 seconds between different recording stations. All time steps were linearly interpolated (as
required) to provide all ground motion accelerograms with a 0.01 second consistent time step
maintaining duration characteristics of the original ground motion. Detailed pre-processing notes
are provided in Table A-4 of Appendix A. It was reasonable not to rotate ground motions to fault
normal and fault parallel directions because no target scenarios were within five km of the site;
however, directionality effects still may impact the results; directionality effects have been left for
future study.
8. Scaling of ground motions
Linear scaling over scenario-specific period ranges of interest was carried out. First, the geometric
mean of bi-directional ground motions was calculated. Next, individual geometric means were
linearly scaled by one divided by the average such that the record specific average was equal to
the response spectrum average. Finally, the code limit of having the mean of geometric means not
fall below 90 percent of the target spectrum was applied as a suite scale factor. The procedures of
each scenario scaling are provided in Figure 5-14, Figure 5-15, and Figure 5-16 and the final mean
spectrum results are presented in Figure 6-13.
82
Crustal - Orthogonal pairs of response spectrums
3.2
Darfield - H1
Darfield - H2
Christchurch - H1
2.4
Christchurch - H2
Imperial Valley - H1
S (g)
1.6 Imperial Valley - H2
Tottori - H1
Tottori - H2
0.8
Parkfield - H1
Parkfield - H2
Target spectrum
0.0
0.0 0.5 1.0 1.5 2.0
T (s)
Unscaled geometric Individually scaled Suite scaled

means over T,rs,i over T,rs,i over T,rs,i
3.2 3.2 3.2
Individual Suite
scale factors: factor:
2.4 2.4 2.4 1.04
0.88 - 1.69
S (g)
S (g)
S (g)
1.6 1.6 1.6
0.8 0.8 0.8
0.0 0.0 0.0

0.0 0.2 0.4 0.0 0.2 0.4 0.0 0.2 0.4
T (s) T (s) T (s)
Darfield Christchurch Imperial Valley Tottori

Parkfield Target spectrum Average
Figure 5-14 Crustal suite
83
In-slab - Orthogonal pairs of response spectrums
Tarapaca_Pica - H1
3.2
Tarapaca_Pica - H2
Tarapaca_Idiem - H1
2.4 Tarapaca_Idiem - H2
Obihiro - H1
Obihiro - H2
S (g)
1.6
Nisqually_DNR - H1
Nisqually_DNR - H2
0.8 Nisqually_Crown - H1
Nisqually_Crown - H2
Target spectrum
0
0.0 1.0 2.0
T (s)
Unscaled geometric Individually scaled Suite scaled

means over T,rs,i over T,rs,i over T,rs,i
3.2 3.2 3.2

Individual Suite
scale factor: 1.63
factors: 0.88
2.4 2.4 - 2.64 2.4
S (g)
S (g)
S (g)
1.6 1.6 1.6
0.8 0.8 0.8
0 0 0
0.0 0.4 0.8 0.0 0.4 0.8 0.0 0.4 0.8
T (s) T (s) T (s)
Tarapaca_Pica Tarapaca_Idiem Obihiro Nisqually_DNR
Nisqually_Crown Average Target spectrum
Figure 5-15 In-slab suite
84
Interface - Orthogonal pairs of response spectrums
3.2
Tohoku_H10 - H1
Tohoku_H10 - H2
Maule_Conc - H1
2.4
Maule_Conc - H2
Maule_CCSP - H1
1.6 Maule_CCSP - H2
S (g)
Hokkaido_T - H1
Hokkaido_T - H2
0.8 Hokkaido_K - H1
Hokkaido_K - H2
Target spectrum
0
0.00 1.00 2.00 3.00 4.00
T (s)
Geometric mean Individually scaled Suite scaled

unscaled over T,rs,i over T,rs,i over T,rs,i
3.2 3.2 3.2
Individual Suite
scale factors: factor:
2.4 2.4 2.4 1.24
1.05 - 1.80
S (g)
S (g)
S (g)
1.6 1.6 1.6
0.8 0.8 0.8
0.0 0.0 0.0

0.30 0.90 1.50 0.30 0.90 1.50 0.30 0.90 1.50
T (s) T (s) T (s)
Tohoku_H10 Maule_Conc Maule_CCSP Hokkaido_T

Hokkaido_K Target spectrum Average
Figure 5-16 Interface suite
85
9. Application of ground motion pairs
While 4.1.8.8.(1) of the 2015 NBCC allows for a uni-directional analysis in each of the two
primary orthogonal axis with orthogonal seismic force resisting systems, a simultaneous bi-
directional analysis was instead selected. The 100:30 rule was applied such that each pair of
orthogonal horizontal ground motions was simultaneously applied with 100 percent of one motion
and 30% of the orthogonal motion in orthogonal directions. Next, each was rotated by 90 ° and the
process was repeated. Vertical components were not included in the analysis. A simultaneous bi-
directional or tri-directional analysis is more appropriate than a uni-directional analysis for
response history analysis (Sherstobitoff, 2008). Also, for vertical elements that resist loads in both
directions, such as columns or core walls, the effects of bi-directionality may become significant
(Canadian Commission on Building and Fire Codes National Research Council of Canada, 2015);
therefore, load cases were defined with the 100:30 rule for orthogonal pairs. This led to 60 load
cases for the three scenario-specific suites of five pairs of orthogonal horizontal ground motions,
allowing each of the two orthogonal ground motions to have 100 percent and 30 percent in each
of the two primary orthogonal horizontal directions. Numerically, 3 x 5 x 2 x 2 = 60 load cases or
30 load cases for each of the two primary orthogonal directions. A sample of the notation for the
load case definition is: CR-1-X-1 where CR indicates crustal, 1 indicates the first ground motion
in the suite (Darfield), X indicated the primary direction (direction of 100 percent), and the 1,
respectively, indicates that the 100 percent component is coming from the H1 component.
5.2.5 Linear response history analysis
For the linear response history analysis, a damping ratio of five percent of critical was assumed
for all modes. The ground motions were scaled up for importance as
𝐼𝑛𝑝𝑢𝑡 𝐺𝑟𝑜𝑢𝑛𝑑 𝑀𝑜𝑡𝑖𝑜𝑛 = (𝐼𝐸 )𝐷𝑒𝑠𝑖𝑔𝑛 𝐺𝑟𝑜𝑢𝑛𝑑 𝑀𝑜𝑡𝑖𝑜𝑛 (5-25)
The initial scale factor was Importance (IE) multiplied by gravity (g) in terms of mm/s2.
Numerically, the scale factors for input accelerations were calculated as (9806)(1.5) = 14,709; due
to a bi-directional analysis being carried out, the simultaneous orthogonal component scale factor
was further reduced as (0.3)(14,709)=4,413. Note that the individual and suite factors were already
applied during the selection and scaling of ground motions procedure. Back scaling was required
for drift. Simultaneous bi-directional pairs were run with the 100:30 rule and the governing
86
response was identified from each pair having either H1 or H2 as primary orthogonal component
(where H1 and H2 are arbitrary orthogonal components of ground motion) as the primary
contributor. Governing combinations were identified and plotted as shown in Figure 5-17 (a), (c),
and (e), and (b), (d), and (f) for x, and y-directions, respectively.
Crustal Crustal
12 12
8
Storey
8
Storey
4 4
0 0
0 0.01 0.02 0.03 0 0.01 0.02 0.03
Drift Drift
RH-CR-1-X-2 RH-CR-2-X-1 RH-CR-1-Y-2 RH-CR-2-Y-1
RH-CR-3-X-1 RH-CR-4-X-1 RH-CR-3-Y-1 RH-CR-4-Y-1
RH-CR-5-X-2 Crustal mean RH-CR-5-Y-2 Crustal mean
Post disaster limit Post disaster limit
(a) (b)
In-slab In-slab
12 12
8 8
Storey
Storey
4 4
0 0
0 0.01 0.02 0.03 0 0.01 0.02 0.03
Drift Drift
RH-IS-1-X-1 RH-IS-2-X-2 RH-IS-1-Y-1 RH-IS-2-Y-2
RH-IS-3-X-1 RH-IS-4-X-1 RH-IS-3-Y-1 RH-IS-4-Y-1
RH-IS-5-X-1 In-slab mean RH-IS-5-Y-1 In-slab mean
(c) (d)
Interface Interface
12 12
Storey
8 8
Storey
4 4
0 0
0 0.01 0.02 0.03 0 0.01 0.02 0.03
Drift Drift
RH-IN-1-X-1 RH-IN-2-X-1 RH-IN-1-Y-1 RH-IN-2-Y-1
RH-IN-3-X-1 RH-IN-4-X-1 RH-IN-3-Y-1 RH-IN-4-Y-2
RH-IN-5-X-1 Interface mean RH-IN-5-Y-1 Interface mean
(e) (f)
Figure 5-17 Governing maximum drift orthogonal pairs with suite average (a), (c), and (e) x-
direction; (b), (d), (f) y-direction
87
Other global response parameters of storey shears and overturning were averaged in the same way;
those results, along with the scenario drift combinations are presented in Chapter Six.
5.2.6 Non-linear response history analysis
Non-linear response history analysis was carried out much the same as the previous linear response
history analysis except with overstrength, non-linearities, damping, and load cases updated.
Overstrength was accounted for by increasing the modulus of elasticity properties by 20 percent
as determined in the literature review section. Strain hardening was included into the reinforcing
model as per the default for A615Gr60 rebar.
The inherent damping was set to 2.5 percent to be more aligned with experimentally measured
values (see the literature review section 2.3.6 Damping), less than the maximum allowable of three
percent under the 2015 NBCC, and greater than the minimum value recommended by the LATBC
(2017). For further explanation of the damping selection.
Fast non-linear analysis (FNA) was used because it saves computational effort and time by
separating the linear and non-linear components in the equation of motion. A Ritz modal case was
defined for use with the FNA.
Dead and live loads were applied as linear static cases. A ramp function was defined with a rise
time of seven seconds and a uniform time of seven seconds. This was to allow the non-seismic
loads to be applied as fast non-linear analysis (FNA) case and then for the time history case to start
after the gravity loads were applied. Damping for the ramp function was set to 99 percent of critical
to allow the gravity loads to be applied and quickly stabilize.
The non-linear hinges included wall fiber hinges in all core walls and coupling beam hinges at
each end of coupling beams in the coupled and double coupled core walls. Cantilever walls exist
on gridline C, F, and H, while double coupled walls exist on grid line four and seven, and single
coupled walls exist on gridline five and six. Note that the single coupled walls were added during
design iterations with the modal response spectrum analysis. Timber connection non-linearities
were omitted from the model to improve computational efficiency; these were previously shown
to provide negligible lateral resistance compared with the core (see Section 4.2.2 ESFP). For the
88
concrete core, the non-linearity was captured as described in the review portion using both fiber
model wall sections for the lower walls plastic hinge region and flexural hinges at coupling beams.
Both hinges were automatically generated by ETABS based on ASCE-41-13 which is the relevant
U.S. standard for seismic evaluation including aspects of performance-based design. Hinge
reinforcing was automatically determined from the design carried out in ETABS. Figure 5-18 (a)
shows the general arrangement in plan, (b) shows each basic type of core wall including coupled,
double-coupled, and cantilever. Run times for the analysis varied by ground motion from about 30
minutes to over eight hours for the more intense, long duration motions. Post processing of global
response parameters, including drift, storey shear and overturning moment, was completed in MS
Excel.
(a) (b)
Figure 5-18 General arrangement (a) plan view with core walls and gravity framing, and (b)
elevations of basic coupled, double-coupled, and cantilever core walls
89
6 Results
Results of experimental verifications and the extended study are presented.
Results of the one, two, and three-dimensional experimental verifications are presented.
The results of best fitting hysteretic connection models are provided in Figure 6-1. Figure 6-1 (a)
and (b) show the beam to core connection for each axis of rotation while (c) and (d) show the
beam-column and column-base connections, respectively. Aside from the less stiff reloading
stiffness, the calibrated connection models provide a good visual trace to the experimental data.
The strong axis of the beam to core connection Figure 6-1 (a) yields at around 25 kNm whereas
the weak-axis (b) yields at around 7 kNm. With the same beam depth and connection arrangement,
the beam to column connection (c) yields slightly lower than the connection to the core. This can
be explained by the crushing of timber column in the perpendicular to grain orientation. The
column to base connection (d) has a limited effective depth which provides yielding at similar
levels to the weak axis beam connection.
Considering link rotational demand from ETABS during the extended study, a maximum rotation
of 0.00181 radians was found. This was during elastic analysis. The rotational demand must
therefore be scaled by (RdRo/IE) equals (3.5*1.7/1.5)(0.00181) = 0.0072 radians as a maximum
inelastic rotational demand. Checking the calibrated connection results, it can be seen that 0.0072
radians is less than both, beam-column at 0.01 radians to yield and beam-core at 0.05 radians to
yield, connection capacities which makes the timber gravity system suitable for inclusion as a
gravity load carrying system.
90
Core to beam - Strong axis Core to beam - Weak axis
30 10
15 5
Moment (kNm)
Moment (kNm)
0 0
-15 -5
-30 -10
-0.05 -0.025 0 0.025 0.05 -0.12 -0.06 0 0.06 0.12
Deformation angle (rad) Deformation angle (rad)
BX_C1 BX_C2 Pivot BY_C1 BY_C2 Pivot
(a) (b)
Beam to column Column to base
30 6
15 3
Moment (kNm)
Moment (kNm)
0 0
-15 -3
-30 -6
-0.05 -0.025 0 0.025 0.05 -0.04 -0.02 0 0.02 0.04
Deformation angle (rad) Deformation angle (rad)
CB_C CB_C Pivot CX_C1 CX_C2 Pivot
(c) (d)
Figure 6-1 Calibrated Pivot hysteretic connection models (a) concrete core to timber beam –
strong axis, (b) concrete core to timber beam – weak axis, (c) timber beam to timber column, and
(d) timber column to base.
Table 6-1 summarizes results of the connection calibration obtained from the connections used in
the 3D timber concrete hybrid model. The backbone parameters are typical between typical
hysteretic models and can be used for various numerical connection models, besides just Pivot;
whereas the non-linear parameters are specific to the Pivot model.
91
Table 6-1 Pivot connection parameters for tension bolt moment resisting connections
Connection type
Parameters Core to beam – Core to beam – Beam to column Column to base
strong axis weak axis
Positive Backbone
M1 [kNm] 24 8.8 26 3.6
Θ1 [radians] 0.009 0.16 0.065 0.003
M2 [kNm] 27 8.3 25 5.7
Θ2 [radians] 0.015 0.11 0.025 0.02
M3 [kNm] 28 7 20 6.5
Θ3 [radians] 0.020 0.057 0.02 0.12
M4 [kNm] 30 2 18 5
Θ4 [radians] 0.032 0.01 0.006 0.15
Negative Backbone
M-4 [kNm] -30 -2 -18 -5
Θ-4 [radians] -0.032 -0.01 -0.006 -0.15
M-3 [kNm] -28 -7 -20 -6.5
Θ-3 [radians] -0.020 -0.057 -0.02 -0.12
M-2 [kNm] -27 -8.3 -25 -5.7
Θ-2 [radians] -0.015 -0.11 -0.025 -0.02
M-1 [kNm] 24 -8.8 -26 -3.6
Θ-1 [radians] -0.009 -0.16 -0.065 -0.003
Stiffness [kNm/radian] 2667 160 2200 140
Non-Linear Parameters
α1 2.8 2.8 2.8 2.8
α2 2.8 2.8 2.8 2.8
β1 0.02 0.1 0.1 0.1
β2 0.02 0.1 0.1 0.1
η 0 0 0 0
92
Table 6-2 provides properties of calibrated connections from the two-dimensional study. The
stiffnesses provided are valid for the specific timber connections studied and do not need to be
used exclusively in the portal frame arrangement.
Table 6-2 Pivot parameters for lagscrewbolt connections
Parameters Beam Connection Column Connection

HTA 400 HTA 500 HTA 600 HCB 4 HCB 5 HCB 8
Positive Backbone
M1 [kNm] 52 50 80 10 25 25
Θ1 [radians] 0.006 0.0025 0.004 0.0045 0.008 0.006
M2 [kNm] 68 80 110 45 55 75
Θ2 [radians] 0.015 0.008 0.008 0.02 0.025 0.035
M3 [kNm] 72 115 110 55 70 95
Θ3 [radians] 0.029 0.02 0.02 0.032 0.045 0.065
M4 [kNm] 30 35 1 35 50 65
Θ4 [radians] 0.065 0.08 0.063 0.046 0.063 0.08
Negative Backbone
M-4 [kNm] -30 -35 -1 -35 -50 -65
Θ-4 [radians] -0.065 -0.08 -0.063 -0.046 -0.063 -0.08
M-3 [kNm] -72 -35 -110 -55 -70 -95
Θ-3 [radians] -0.029 -0.02 -0.02 -0.032 -0.045 -0.065
M-2 [kNm] -68 -115 -110 -45 -55 -75
Θ-2 [radians] -0.015 -0.008 -0.008 -0.02 -0.025 -0.035
M-1 [kNm] -52 -80 -80 10 -25 -25
Θ-1 [radians] -0.006 -0.0025 -0.004 -0.0045 -0.008 -0.006
Semirigidity [kNm/radian] 4,370 8,900 16,000 2,600 2,800 3,500
Non-Linear Parameters
α1 2.8 2.8 2.8 2.8 2.8 2.8
α2 2.8 2.8 2.8 2.8 2.8 2.8
β1 0.1 0.1 0.1 0.1 0.1 0.1
β2 0.1 0.1 0.1 0.1 0.1 0.1
η 0 0 0 0 0 0
93
Pushover validations for the different types of column to base and beam to column connections
are shown in Figure 6-2 (a) and (b), respectively. The component pushover curves show the
numerical model fitting well to the experimental results. Next, the results of the portal frames with
400 mm and 500 mm deep beams are shown in Figure 6-3 (a), and (b), respectively.
40 40
Load (kN)
Load (kN)
20 20
0 0
0 40 80 120 160 0 40 80 120 160
Deformation (mm) Deformation (mm)
HTA 400-1 HTA 400-2
HCB4-1 HCB4-2
HCB4-3 HCB4 - ETABS HTA 400-3 HTA 500-1
HCB5-1 HCB5-2 HTA 500-2 HTA 500-3
HCB5-3 HCB5 - ETABS HTA 600-1 HTA 600-2
HCB8-1 HCB8-2 HTA 600-3 HTA 400 - ETABS
HCB8-3 HCB8 - ETABS HTA 500 - ETABS HTA 600 - ETABS
(a) (b)
Figure 6-2 Connection static pushover curves matched to experimental data (a) column to base,
and (b) beam to column
150 150
Load (kN)
100 100
Load (kN)
50 50
0 0
0 50 100 150 200 250 0 50 100 150 200 250
Deformation (mm) Deformation (mm)
HR 400-1 HR 500-1
HR 400-2 HR 500-2
HR 400-3 HR 500-3
HR 400 ETABS - PIVOT HR 500 ETABS - PIVOT
(a) (b)
Figure 6-3 Frame static pushover curves matched to experimental data (a) 400mm beam depth
frame, and (b) 500mm beam depth frame
94
Results from the non-linear static pushover analysis of the portal frames fitted the experimental
data well. This was achieved through iterations of connection stiffness at the connection level, then
application of the validated hysteretic models to the portal frame models. Validation of the
calibrated connection models to the portal frame models allowed for further confidence in the
connection parameters presented in Table 6-2 as well as verified the modelling techniques used to
develop the two-dimensional non-linear frame model.
Best matching of floor acceleration and displacement are shown in Figure 6-4 and Figure 6-5,
respectively.
3000
Acceleration (gal)
1500
0
-1500
-3000 -2310.03 -2517.06
0 25 50 75 100
Time (sec)
Figure 6-4 Best fit of maximum floor acceleration
Figure 6-4 shows peak values within eight percent, (2517-2310)/2517=0.08; however, the phase
and alignment of peaks is not well fitted. Similarly, Figure 6-5 shows overall maximum floor
displacements within 11 percent (15.95-17.80)/15.95=0.11 but not well fitted phase and alignment.
The best simultaneous matching of floor acceleration and peak displacement occurred with
damping set to around 2.2 percent and diaphragm stiffness was set to 12,000 MPa. The best
individual matching of floor acceleration was achieved with very low damping of about 0.5
percent; however, it was apparent that this was not correct due to the rapid oscillations of the under-
damped system. For individual matching of maximum displacement, diaphragm stiffness was
reduced; however, the floor accelerations became far lower than the experimental values;
therefore, the results were not considered representative of the real system. Another aspect that
was not well matched was the experimental data which showed the maximum displacement
occurring at the first storey. All numerical simulations in this study predicted maximum
displacement in the upper storey. Perhaps some additional non-linearity occurred in the experiment
95
which was not accounted for in the numerical model; investigation of this has been left for future
study.
40
15.95 mm 17.80mm
Displacement (mm)
20
0
-20
-40
0 25 50 75 100
Time (sec)
Figure 6-5 Best fit of maximum floor displacement
6.2 Extended study
Results from each analysis type and from the selection and scaling of ground motions are presented
for the extended study of the 12-storey structure.
6.2.1 Equivalent static force
Final maximum drifts below the code specified one percent drift limit for post disaster structures
was achieved through design iteration, including increasing thickness of coupling beams to 850mm
(as shown in Figure 6-9), increasing concrete strength to 60 MPa, increasing core wall thickness
as shown in Figure 6-9 and finally, the addition of two coupled blade walls, also as shown in Figure
6-9. The final drifts from equivalent static force procedure are plotted in Figure 6-6. As with the
preliminary design, the softer coupled system in the x-direction is governing the drift response at
approximately 0.90 percent while the stiffer, cantilevered shear wall system (in the y-direction) is
at 0.85 percent. For both directions, the maximum interstorey drift occurs just beyond the mid-
height of the structure. The increase in concrete volume required to achieve passing design for the
coupling beams and shear wall piers in the coupled-direction was significant. The increase over
the base structure results presented in Figure 4-11 on page 47 of this Thesis can be explained by
the increase in strength demand and reduced drift allowance due to high importance; as well as,
the amplified response spectrum due to soft soil. Further, the increase in self weight of the core
due had a feedback effect that further increased the design requirements.
96
Drift
12
Storey
4
0
0 0.004 0.008 0.012
Drift
x-direction y-direction Post disaster limit
Figure 6-6 Equivalent static drift
Storey shears plotted in Figure 6-7 show uncoupled (y-direction) results with higher shear forces
due to the stiffer system compared to the coupled (x-direction) system. Similarly, higher
overturning is shown for the stiffer (un-coupled) system in Figure 6-8.
Storey shear
12
8
Storey
0
-20,000 -15,000 -10,000 -5,000 0
Shear (kN)
Figure 6-7 Equivalent static storey shear
Overturning moment
12
8
Storey
0
-400,000 -300,000 -200,000 -100,000 0
Overturning (kNm)
Figure 6-8 Equivalent static overturning moment
97
Final reinforced concrete core section and elevation reinforcing, as developed with ETABS, are
provided in Figure 6-9. The results show that it was necessary to add two blade walls to maintain
drifts of less than one percent. This design option was selected in order to not change the overall
core dimensions and it was seen as a way to achieve a passing design without disrupting the overall
flow excessively. Alternate design options were not investigated but may prove to be more
practical. The thicker walls in the coupled direction are aligned with findings of other researchers
(CSCE, 2008) in that shear demand in coupling beams is critical and that dimensional iterations
are often required to reduce demand and/or increase capacity.
98
(a)
(b)
Figure 6-9 Concrete core (a) section, and (b) elevation
99
6.2.2 Response spectrum
Final drift plots are provided in Figure 6-10; note that the x-direction appears to slightly exceed
the limit; however, within the required significant figures, the drift is at the limit and acceptable.
Drift
12
Storey 8
4
0
0 0.004 0.008 0.012
Drift
x-direction y-direction Post disaster limit
Figure 6-10 Response spectrum drift
Storey shears plotted in Figure 6-11 align well with the ESFP results maintaining higher relative
demand of the stiffer y-direction system.
Storey shears
12
Storey
8
4
0
0 5000 10000 15000 20000
Shear (kN)
Figure 6-11 Response spectrum storey shear
Overturning plotted in Figure 6-12 shows lower overall demand compared to the ESFP
overturning, indicating influence of higher modes in the vertical distribution of seismic loads.
Overturning moment
12
8
Storey
4
0
-400,000 -300,000 -200,000 -100,000 0
Overturning (kNm)
Figure 6-12 Response spectrum overturning moment
100
6.2.3 Selection and scaling of ground motions
Results of the selection and scaling of ground motions are provided in Table 6-3 and the scaled
mean spectra is provided in Figure 6-13.
Table 6-3 Final selected suites of ground motions

Individual Suite
R Vs30 scale scale
Scenario Earthquake Year Site Mw (km) (m/s) factor factor
Darfield, NZ 2010 Resthaven 7.0 19 141 1.64
Christchurch, NZ 2011 Resthaven 6.2 5 141 1.38
Crustal Imperial Valley, CA 1979 E.C.A.-3 6.5 13 163 1.69 1.04
Tottori, Japan 2010 TTR008 6.6 7 139 1.48
Parkfield, CA 2004 P.F.Z.-1 6.0 3 178 0.88
Tarapaca, Chile 2005 Pica 7.8 125 492 0.88
Tarapaca, Chile 2005 Idiem 7.8 144 386 2.61
In-slab Obihiro, Japan 2013 TKCH07 6.9 107 140 1.53 1.63
Nisqually, WA 2001 Crown 6.8 78 285 2.07
Nisqually, WA 2001 DNR 6.8 56 300 2.64
Tohoku, Japan 2011 IBRH10 9.0 338 144 1.63
Maule, Chile 2010 Conc 8.8 69 223 1.75
Interface Maule, Chile 2010 CCSP 8.8 70 236 1.8 1.24
Hokkaido, Japan 2003 TKCH07 8.0 125 140 1.05
Hokkaido, Japan 2003 KSRH02 8.0 150 219 1.37
Scaled mean spectra over Trs

2.4
1.8
S (g)
1.2
0.6
0.9S(T) 0.9S(T) 0.9S(T)
0
0 0.4 0.8 1.2 1.6 2
T (s)
Crustal 90% target In-slab 90% target Interface 90% target
Crustal mean In-slab mean Interface mean
Crustal mean within T,rs,i In-slab mean within T,rs,i Interface mean within T,rs,i
Figure 6-13 Scaled mean spectra
101
Limited historical records made finding high quality matches difficult. The ground motions, along
with their scaled mean spectrum results are aligned with findings of Tremblay et al. (2015).
Increased demand is evident due to short period over scaling resulting from the default linear
scaling method of the guideline.
6.2.4 Linear response history
Scenario-specific average maximum results from the simultaneous bi-directional linear response
history analysis are provided. Figure 6-14 provides interstorey drift results which are not passing
for the post-disaster limit; however, they are well below high and normal importance limits. The
drifts exceeding the limit was not a surprise as the scaled mean spectrum for each suite of ground
motions is above the target spectrum. The response history analysis is beyond the code requirement
and serves to better predict and understand structural response; therefore, design iterations were
not carried out.
Suite average maximum drift in x-direction
12
8
Storey
4
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Drift
(a)
Suite average maximum drift in y-direction
12
8
Storey
4
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Drift
Crustal mean In-slab mean Interface mean
Post disaster limit High importance limit Normal importance limit
(b)
Figure 6-14 Linear response history suite average maximum drift (a) x-direction, and (b) y-
direction
102
Scenario-specific shear demand is provided in Figure 6-15. Both subduction-type hazards (in-slab
and interface) govern the demand at all storeys. Similarly, shows overturning dominated by
subduction hazards. Figure 6-16 shows both major and minor bending due to the 100:30 load
application. It can be seen that the minor subduction overturning is approximately equal to the
crustal major overturning.
The results demonstrate a benefit of response history analysis; that is the unique positive and
negative time varying response to allow more realistic load combinations to be generated. Further,
the simultaneous bi-directional response history analysis allowed determination of simultaneous
minor axis loading concurrent with the major axis loading.
Suite average maximum storey shears in x-direction

12
8
Storey
0
-15,000 -10,000 -5,000 0 5,000 10,000 15,000
Shear (kN)
(a)
Suite average maximum storey shears in y-direction
12
8
Storey
0
-15,000 -10,000 -5,000 0 5,000 10,000 15,000
Shear (kN)
Crustal average Crustal average In-slab average
(b)
Figure 6-15 Linear response history suite average maximum storey shear (a) x-direction, and (b)
y-direction
103
Suite average maximum overturning moment in primary x-direction
12
8
Storey
0
-320,000 -240,000 -160,000 -80,000 0 80,000 160,000 240,000 320,000
Overturning (kNm)
(a)
Suite average maximum overturning moment in primary y-direction
12
8
Storey
0
-320,000 -240,000 -160,000 -80,000 0 80,000 160,000 240,000 320,000
Overturning (kNm)
Crustal average major negative Crustal average major posative
Crustal average minor negative Crustal average minor posative
In-slab average major negative In-slab average major posative
In-slab average minor negative In-slab average minor posative
Interface average major negative Interface average major posative
Interface average minor negative Interface average minor posative
(b)
Figure 6-16 Linear response history suite average maximum major and minor overturning
moment (a) x-direction, and (b) y-direction
Member forces for the base of a lower storey pier are provided from a linear response history
analysis which used preliminary ground motions. The results are meant to convey load
combinations rather than be used for a specific design. Tension and compression are uniquely
handled for load cases (due to the non-uniformity of reinforced concrete materials and sections
behavior under tensile and compressive loading). Also, it is shown that the maximum axial load is
not concurrent with the maximum shear and bending values. First, the results were averaged then
each of the averaged load cases for x- and y-direction inputs were applied (as a set) either entirely
negative or entirely positive. The averaged output results and resulting four load cases (for seismic
loading) are shown in Table 6-4.
104
Table 6-4 Member design forces from preliminary ground motions
Axial Shear Shear Torsion Moment Moment
Case (kN) (kN) (kN) (kNm) (kNm) (kNm)
Case 1 and 2 (+/-) 320 1701 0 24 0 29668

Case 3 and 4 (+/-) 8719 285 42 25 101 1224
Case one (1) and two (2) represent input load functions where the primary function is in-plane
stiffness. Case three (3) and four (4) represent out-of-plane loading (relative to the SFRS primary
direction). The large axial forces are due to the flanging effect that occurs when in-plane SFRS
walls are loaded in their weak axis but they are fully shear connected to the in-plane walls. The
out-of-plane elements provide negligible base moment resistance but they resist significant
moment by coupling effects between opposite flanges of an in-plane SFRS component. This
member design force example illustrates a benefit of a linear response history analysis rather than
a linear modal response spectrum analysis as it allows signs to be preserved and timewise variation
in loads which allows for more realistic load combinations. All further post-processing of results
and design of elements was left for future study.
6.2.5 Non-linear response history
Scenario-specific average maximum results from the simultaneous bi-directional non-linear

response history (NL-RH) analysis are provided. Figure 6-17 provides interstorey drift results
which again show the subduction-type seismic events dominating the response. Drifts are
reasonably aligned with the linear response history results which is as should be expected
according to National Research Council (2015). As typical, the interface and in-slab dominate the
response; however, the dynamic response of coupled versus uncoupled systems were uniquely
excited in the non-linear analysis which shows them alternately dominating the drift response.
Storey shears for each primary direction are presented in Figure 6-18. Overturning moment results
in Figure 6-19 (a) show typical results of subduction events dominating the response in the coupled
direction; however, in the cantilever (y) direction, the response to crustal dominates over in-slab
scenarios in Figure 6-19 (b). Because this response was not evident in any linear analysis
conducted nor in the cantilever (y) direction, the coupling beam non-linearity may have
contributed to the enhanced response to the crustal suite of ground motions. The non-linear results
105
show excessively high shear and overturning results; however, the drift results are in-line with the
other analysis method results. The high shear and overturning values indicate that insufficient non-
linear response is being achieved and force demand is not being reduced to the extent that the
structural system has been designed. This may be a result of the strict drift limitation of one percent
imposed by the high importance classification of the structure which in turn made the structure
unable to achieve ductility in the anticipated range or it may be that hinge properties need to be
refined. Further and detailed investigation of non-linear hinge properties used at coupling beams
and in wall segments has been left for future study but should be investigated further, as discussed
in the conclusion chapter of this thesis.
Suite average maximum drift in x-direction

12
8
Storey
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Drift
(a)
Suite average maximum drift in y-direction
12
8
Storey
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Drift
Crustal In-slab Interface
Post disaster limit High importance limit Normal importance limit
(b)
Figure 6-17 Non-linear response history suite average maximum drift (a) x-direction, and (b) y-
direction
106
Suite average maximum storey shears - Primary x-direction loading
12
8
Storey
0
-150,000 -100,000 -50,000 0 50,000 100,000 150,000
Shear (kN)
(a)
Suite average maximum storey shears - Primary y-direction loading
12
8
Storey
0
-150,000 -100,000 -50,000 0 50,000 100,000 150,000
Shear (kN)
Interface average major negative Interface average major posative
Interface average minor negative Interface average minor posative
(b)
Figure 6-18 Non-linear response history suite average maximum storey shear (a) x-direction,
and (b) y-direction
107
Suite average maximum overturning moment - Primary x-direction loading
12
8
Storey
0
-4,000,000 -2,000,000 0 2,000,000 4,000,000
Overturning (kNm)
(a)
Suite average maximum overturning moment - Primary y-direction loading
12
8
Storey
0
-4,000,000 -2,000,000 0 2,000,000 4,000,000
Overturning (kNm)
Interfaceaverage major negative Interfaceaverage major posative
Interface average minor negative Interfaceaverage minor posative
(b)
Figure 6-19 Non-linear response history suite average maximum overturning moment (a) x-
direction, and (b) y-direction
108
7 Conclusions
Conclusions from the experimental validations and extended study are provided.
Conclusions from the one, two and three-dimensional studies are presented.
Hysteresis matching with the Pivot model was used to validate non-linear parameters of specific
timber and timber-concrete hybrid connections. The Pivot model provided traces that fit generally
well; however, limitations are noted along with the contributions of this research.
The key limitation of this research and the Pivot model was that accurate initial reloading stiffness
for the timber and concrete-timber hybrid connections could not be realized. Regardless of non-
linear parameter definitions, the initial reloading stiffness was less stiff in the Pivot model
compared to the experimental. The Pivot model restricted initial reloading stiffness to be more
gradual, gaining strength steadily from the axis of zero displacement. The pivot point (at the axis
of zero displacement) was often not representative of actual connection performance. This can be
explained by the wood crushing due to previous excursions and subsequent lack of stiffness until
new contact is realized. This results in underprediction of the energy dissipation per hysteretic loop
in the Pivot model (based on the area enclosed in the hysteretic loop) compared to the energy
dissipation occurring in reverse cyclic load tests.
The second limitation of this research is along unloading stiffness trace which degraded more in
the Pivot model compared to that of the reverse cyclic testing. This was only slightly apparent even
near four times the yield rotation. The slight loss of unloading stiffness does not appear to be
significant; however, quantification of this limitation was beyond the scope of this study.
The first contribution of this research is the observation that non-linear parameters of the Pivot
hysteresis model did not vary significantly between all connections calibrated. This indicates that
basic assignment of the non-linear parameters may be valid for similar connection types which
109
would leave the user to only update backbone curve parameters, simplifying the utility of the
connection model.
The second contribution of this research is the recommendation to use the Pivot model for timber
connections in timber and concrete-timber hybrid structures until such a time that an improved
model is provided.
The third contribution of this research is the provision of validated hysteretic connection models
for the four connection types that were tested. It is recommended to use the linear and non-linear
parameters of the calibrated connections as provided in the results section of this study for timber
and concrete-timber hybrid connections the same as those validated.
The fourth contribution of this research is the application of hysteretic validation in both major
and minor axis flexure for the beam to core connection. This type of validation will foster use of
the validated connection model in a three-dimensional system. Where connections are not
validated for weak axis flexure, connection model parameters would be left as free to rotate and
not provide and stiffness to the system nor would the effects of hysteretic energy dissipation be
captured due to weak axis flexure.
The fifth contribution of this research was recognition of the fact that improved unloading and
reloading stiffness capabilities of the Pivot model in ETABS would improve accuracy in terms of
energy dissipation and trace-characteristics of the hysteresis. The best fitting model that was
currently available was Pivot. If the model could maintain very little stiffness until the rotation
from the previous cycle was obtained, a better fit may be achieved.
Expanding on the fifth and final contribution is a recommendation for inclusion of an additional
Pivot parameter to shift the positive and negative pivot points respectively in the opposite
directions along the rotational (x) axis of the hysteretic, moment-rotation plot to improve the Pivot
connection model. The additional parameter could allow for definition of a factor, normalized
between zero and one (0-1), where zero would provide a result exactly as the present Pivot model
(where reloading stiffness begins to ramp up as soon as rotation becomes non-zero. When the
proposed parameter approaches one, the ramping up or reloading stiffness would hold until the
previous cycle rotation was achieved. The additional parameter could be defined for the negative
110
side of the model with another separate parameter, allowing for unique behaviour when subjected
to positive compared to negative flexural stress.
This one-dimensional study was based on four connection types tested not less than twice each;
increased sample size and connection type testing, as well as increased number of hysteretic
models validated (beyond just Pivot) would increase applicability and allow more broad
conclusions to be drawn.
Non-linear static pushover analysis and hysteresis matching were used to validate the portal frame
connections and portal frame global drift response which demonstrated good fit to experimental
results.
The first limitation of this research is same as in the one-dimensional study, related to the fit of the
Pivot model. Please refer to the one-dimensional study for full explanation related to hysteretic
matching.
The second limitation of this research is that the two-dimensional study does not include three-
dimensional effects; therefore, any influence of minor axis flexure is not captured in the two-
dimensional pushover analysis.
The first contribution of this research is the provision of validated Pivot connection model
parameters for a series of post to beam and post to base connections. These connection models
may be used in timber or concrete-timber hybrid structures.
The second contribution of this research is the capture of non-linear response using non-linear
static pushover analysis on a two-dimensional frame with validated hysteretic connections oriented
in two unique global directions. This shows a practical application of using validated connection
models which is not a new concept; however, the specific connection and frame arrangement has
not previously been demonstrated with the Pivot connection model.
This research was based on three experimental tests of each frame, beam depth and connection
type; increased sample size and increased overall testing scheme would allow wider application of
111
the results. Further parametrization of timber frame types and timber connection methods, with
connections to ensure ductile failure modes, would improve the implementation of semi-rigid
timber portal frames in practice.
Full scale shake table testing was used to validate the three-dimensional numerical model. This
three-dimensional study fostered confidence in extension to the more advanced numerical
modelling required for the extended study. The targeted non-linear response parameters were
captured using a lumped plasticity approach with the Pivot hysteretic connection models
previously validated in the one-dimensional study. Contributions and limitations of the three-
dimensional validation are discussed.
The first and more general limitation of this research is that the full dynamic response of the two-
storey, three-dimensional structure was not fully captured. While great efforts were made to
validate key parameters of non-linear connections, damping, diaphragm stiffness and core
stiffness, the model simply did not provide in-phase response of the shake table results.
A second limitation of this work is that peak displacements of the shake table work occurred in
the first storey; whereas, in the three-dimensional model, the peak displacements occurred in the
upper storey. This may indicate either additional non-linearity in the first storey or a contribution
of higher mode effects to the dynamic response.
The main contribution of this research is that damping was found to fit best at 2.2 percent based
on simultaneous matching of peak floor accelerations and displacement as well as review of the
response history trace for frequency of oscillations. The value of 2.2 percent seems to be an
appropriate estimate (for use with non-linear analysis of hybrid concrete-timber structures)
because it fits well within the range of experimental and recommended values as discussed in this
study.
The secondary contribution of this research is the demonstration of application of validated

hysteretic connection models to a three-dimensional numerical model. Biaxial hysteretic
connection parameters were used which demonstrated a method that could be used by designers
112
to evaluate demand on timber connections, particularly in a torsionally sensitive structure, during
non-linear response history analysis of timber or concrete-timber structures.
Data from a single shake table experiment formed the basis of this study; further numerical
validations of the other parametrized structures could be used to refine modelling aspects and
develop innovative solutions for hybrid concrete timber structures. If appropriate ductility and
overstrength factors could be established for a hybrid timber concrete lateral force resisting system,
then, where feasible, the timber semi-rigidities could be attributed to the seismic force resisting
system.
For future work to improve three-dimensional model validation results, it is recommended to

thoroughly review all aspects of the shake table experiment and compare with numerical model
parameters to identify any potential discrepancies.
7.2 Extended study
Conclusions from the extended study of a 12-storey hybrid structure, including limitations and
contributions, are presented.
The first limitation of the extended study is that only a single structure was studied with only a few
design iterations carried out and foundation system was neglected. Options such as increasing the
concrete core size or adding additional shear walls further from the centre of rigidity could
significantly reduce material volumes. Further, analysis and design of the required foundation
system to allow overall structure performance to be verified should also be included.
The second limitation of this research is that the concrete hinge property definition was not
explored in great detail but instead defined based on current design. The non-linear response
history analysis demonstrated that at the low target drift of one percent, minimal non-linear
behaviour was occurring in the concrete hinges which explains the large forces observed in the
non-linear analysis results compared with those of the linear response history analysis. This may
indicate that when the structure is designed to such stringent drift criteria, there is less non-linearity
occurring than the ductility factor would suggest. Alternately, the large forces may also be due to
incorrect hinge property definition. Iteration of hinge properties was beyond the scope of this
research.
113
The third limitation of this research, realized during the selection and scaling of ground motions
procedure, was the limited availability of historical ground motion records. This well-documented
concern has led to various solutions including developing synthetic ground motions and relaxing
ground motion selection criteria. In this study, the limited availability of historical records was
overcome by relaxing the criteria for either the soil type, magnitude, or distance in order to meet
the guideline criteria of getting multiple unique records; however, this relaxation did not come
without cost. Relaxing the soil criteria to allow stiffer soils contributed to the eventual over scaling
in the short period range due to the increased low frequency content (spectral shape) of stiffer soil
sites. Conversely, excessive relaxing of the magnitude or distance criteria resulted in increased
scale factors (at times beyond recommended values). Potential inclusion of synthetic record was
also considered; however, these were eventually eliminated from final selection due to lack of
guidance for inclusion within suites under the guideline and perceived general preference of
historical records based on the literature review.
The first contribution of this research is the identification of the primary structural components of
the lateral and gravity systems and their interaction within the hybrid structure. The primary lateral
system components include vertical and horizontal elements which are the reinforced concrete
core and the cross laminated timber diaphragms, respectively. The primary gravity system
components include the timber post and beam framing as well as the cross laminated timber floor
panels. Identification of the structural components to be included in the primary lateral system
required checking the contribution of timber frame semi-rigidities to the first mode period as well
as consideration of the impact on the resulting lateral design load requirements. For the impact on
the first mode period, the timber frame semi-rigidities were found not to contribute, causing only
a two percent decrease in the first mode period. According to the national building code of Canada
(2015), if the influence is less than 15 percent, the contribution may be neglected during the lateral
design. For the impact on the lateral force demand, inclusion of the timber frame semi-rigidities
would require reduced ductility and overstrength factors which would increase the design load and
be a detriment on the system; therefore, the timber semi-rigidities were not included in the lateral
system. Once it was determined that the timber frames were not to be included in the lateral system,
their interaction with the gravity system was considered. It was required to check if they could
accommodate movements expected under the design ground motion movement. The timber frames
were found to be able to accommodate the expected lateral movements for the drift limit of one
114
percent as required for the post disaster drift limit imposed on this structure. For future work,
ductility and overstrength factors for hybrid structural systems should be developed to better reflect
the actual design base shear that should be considered for respective systems so that current worst-
case overstrength and ductility factors would not be required when combining lateral system
components from unique classifications of seismic force resisting systems.
The second contribution of the research is the preliminary design of the vertical component of the
lateral system; namely, the concrete core, including cantilever walls and coupled cantilever walls.
Design iterations were required from the initial structure which was based on significantly less
loading. Design iterations resulted in the addition of coupled blade walls to control drift and an
increase in wall thickness and concrete strength primarily to meet strength requirements. Future
work should aim to reduce concrete volumes by increasing effective depth of core walls or;
alternately, changing the concrete core to a lighter, perhaps mass timber core may be suited to
further investigation.
The third contribution of this research is the selection and scaling of ground motions for the soft
soil site of Victoria, B.C. Three suites of five ground motions were selected and scaled for the site
and structure type. This facilitated a response history analysis with inputs, developed in accordance
with the building code guideline, that could be reasonably expected for the site. The specific
ground motions could be used for other structures with similar period ranges of interest, soil type,
and location.
A fourth contribution of this research is identification of three aspects of the guideline for selection
and scaling of ground motions which did not provide adequate guidance to ensure consistent
application. These are 1) definition of the period range of interest, 2) identification of target
scenarios, and 3) the relative importance of ground motion selection criteria - magnitude, distance,
and soil type. The most critical of the three is definition of a period range of interest. This is
because the defined period range of interest directly affects the ground motion scale factors which
impact the resulting structural response. The second aspect is the relative importance or weighting
of magnitude, distance and soil type criteria during the ground motion selection process. Guidance
here would reduce variability in ground motion selection results which would in-turn reduce
variability in predicted structural response. The third area of the guideline identified as needing
attention was in the definition of target scenarios. Definition of target scenarios was attempted
115
from a few perspectives in this research including global mean, global mode, and inspection of
deaggregation plots. The deaggregation data that was used was certainly critical to developing the
scenarios that were developed in this work; however, there was still some judgement required in
determining scenario-specific magnitude and distance targets. Improved guidance on the definition
of target scenarios is seen as a less critical task because the exact target scenarios did not have a
significant impact on ground motion selection due to the limited availability of historical records
and the resulting required relaxation of scenarios for selection. Still, a consistent approach
provided in the guideline would be beneficial and the inclusion of scenario specific means within
deaggregation data would be helpful in achieving consistent target scenario selection.
The final contribution of this research is the provision of four methods of analysis from the building
code to the hybrid concrete-timber structure. These methods allow for comparison of results from
each method as well as provide for confidence in the results of more complex analysis methods.
Further, use of the dynamic analysis methods allows a savings in force demand compared with the
static approach which results in design savings. The response spectrum method allowed for
improved estimation of vertical force distribution within the structure which improves the quality
of the response prediction. The linear response history method provides opportunity to develop
ground motion selection and scaling procedures as well as create load combinations that are not
possible in the response spectrum method. The non-linear response history method is beyond the
code requirement for normal structures but where special studies are required such as in developing
innovative hybrid structural systems, non-linear response history analysis can provide a better
prediction of the structural response. In this structure, the non-linear results appear to need refining
to be reliable as they should produce similar results as the previous analysis methods. This
refinement of non-linear results is left for future study.
116
References
Abeysekera, I. (2015). Dynamic response of tall timber buildings. Institute of Structural

Engineers, (May, Issue 101)
Adebar, P., Mutrie, J., & DeVall, R. H. (2005). Ductility of concrete walls: The Canadian
seismic design provisions 1984 to 2004. Canadian Journal of Civil Engineering, 32, 1124-
1137.
Ancheta, T., Darragh, R., Stewart, J., Seyhan, E., Silva, W., Chiou, B., . . . Donahue, J. (2018).
PEER NGA-West2 database. Retrieved December 10, 2017 from
https://ngawest2.berkeley.edu/
Applied Technology Council. (2009). FEMA 695: Quantification of building seismic performance
factors. ( No. 695). Redwood City, California: FEMA.
Archila, M., & Ventura, C. (2012). (2012). Effect of ground motion directionality on seismic
response of tall buildings. Paper presented at the 15th World Conference of Earthquake
Engineering, Lisboa.
Association of professional engineers and geoscientists of British Columbia (APEGBC). (2015).

Technical and practice bulletin 5 and 6 storey wood frame residential building projects (mid-
rise). Canadian Wood Council.
Association of professional engineers and geoscientists of British Columbia (APEGBC). (2016).

APEGBC professional practice guideline: Sustainability. Volume 1.
Atkinson, G. (2009). Earthquake time histories compatible with the 2005 NBCC uniform hazard
spectrum. Canadian Journal of Civil Engineering, 36, 991-1000.
Atkinson, G., & Macias, M. (2008). Predicted ground motions for great interface earthquakes in
the Cascadia subduction zone. Bulletin of the Seismological Society of America 1-60.
117
Baker, J. W. (2011). Conditional mean spectrum: Tool for ground-motion selection. Journal of
Structural Engineering, 137(3), 322-331. doi:10.1061/(ASCE)ST.1943-541X.0000215
Bartlett, S. (2014). Processing of time histories. Department of Civil and Environmental

Engineering, University of Utah.
Bastias, N., & Montalva, G. (2016). Chile strong ground motion flatfile. Earthquake Spectra,
(Preprint)
Binational Softwood Lumber Council. (2017). In Holt R., Luthi, T. and Dickof, C. (Eds.), Nail-
laminated timber: Canadian design and construction guide v1.1 (1st ed.). Vancouver, BC:
Binational Softwood Lumber Council and Forestry Innovation Investment Ltd.
Boivin, Y., & Paultre, P. (2012). Seismic force demand on ductile reinforced concrete shear walls
subjected to western North American ground motions: Part 1 - parametric study. Canadian
Journal of Civil Engineering, (39), 723-737. doi:10.1139/L2012-043
Boore, D. (2005). On pads and filters: Processing strong-motion data. Bulletin of the Seismological
Society of America, 95(2), 745-750. doi:10.1785/0120040160
Booth, E., & Key, D. (2006). Timber. Earthquake design practice for buildings (second ed., pp.
219). Heron Quay, London: Thomas Telford.
Boroschek, R., & Yanez, F. (2000). Experimental verification of basic analytical assumptions used
in the analysis of structural wall buildings. Engineering Structures, 22(6), 657-669.
Braconi, A., Salvatore, W., Tremblay, R., & Bursi, O. S. (2007). Behaviour and modelling of
partial-strength beam-to-column composite joints for seismic applications. Earthquake
Engineering & Structural Dynamics, 36(1), 142-161. doi:10.1002/eqe.629
Canadian Commission on Building and Fire Codes National Research Council of Canada. (2015).
Structural commentaries (users guide - NBC 2015: Part 4 of division B) NRC - National
Research Council Canada.
118
Canadian Hazards Information Service. (2017). In correspondence from Halchuk, S. (Ed.), Seismic
hazard deaggregation Natural Resources Canada.
Canadian Wood Council and American Wood Council. (2016). User guide: For U.S. design office
10 and Canadian design office 9 Canadian Wood Council.
CESMD. (2018). Strong motion virtual data center (VDC). Retrieved on April 12, 2018 from
http://strongmotioncenter.org/vdc/scripts/about.plx
Chang, W., Shanks, J., Kitamori, A., & Komatsu, K. (2009). The structural behavior of timber
joints subjected to bi-axial bending. Earthquake Engineering & Structural Dynamics, 38(6),
739-757. doi:10.1002/eqe.854
Charney, F. (2015). (2015). Linear response history analysis procedure for the 2015 NEHRP
recommended provisions and for ASCE 7-16. Paper presented at the American Society of Civil
Engineers Structures Congress, Portland, Oregon. 2467-2482.
Computers and Structures Inc. (2015). ETABS Volume 15
Computers and Structures Inc. (2016). Rigid vs. semi-rigid. Retrieved on June 30, 2017 from
https://wiki.csiamerica.com/display/etabs/Rigid+vs.+Semi-rigid+diaphragm
Computers and Structures, I. (2015a). CSI analysis reference manual. Berkeley, California:
Computers and Structures, Inc.
Computers and Structures, I. (2015b). CSI analysis reference manual: For SAP2000, ETABS,
SAFE, and CSIBridge. (No. July).
COSMOS. (2005). Guidelines and recommendations for strong-motion record processing and
commentary. (No. CP-2005/01). California: COSMOS.
CSA - CSA Standards. (2014a). CSA CAN/CSA-A23 Design of concrete structures
CSA - CSA Standards. (2014b). CSA CAN/CSA-O86 Engineering design in wood.
119
CSA - CSA Standards. (2014c). CSA CAN/CSA-S16 Design of steel structures.
CSCE. (2008). (2008). Time history analysis. Paper presented at the Time History Analysis
Presented by CSCE, UBC Vancouver. 1-5.
CSI Inc. Generation of spectrum-compatible time-history functions by modifying a reference time

series in the frequency/time domain. Retrieved from
http://docs.csiamerica.com/helpfiles/etabs/Menus/Define/Functions/Time_History_Function
s/Time_History_Matched_to_Response_Spectrum.htm
Daneshvar, P., Bouaanani, N., Goda, K., & Atknison, G. (2016). Damping reduction factors for
crustal, in-slab, and interface earthquakes characterizing seismic hazard in southwestern
British Columbia, Canada. Earthquake Spectra, 32(1), 45-74.
Dawood, H., Rodriguez-Marek, A., Bayless, J., Goulet, C., & Thompson, E. (2016). A flatfile for
the KiK-net database processed using an automated protocol. Earthquake Spectra, 32(2),
1281-1309.
Dehghani, M., & Tremblay, R. (2016). Robust period-independent ground motion selection and
scaling for effective seismic design and assessment. Journal of Earthquake Engineering,
20(2), 185-218. doi:10.1080/13632469.2015.1051635
Departmento de Ingenieria Civil. (2018). Terremotos de Chile. Retrieved on April 15, 2018 from
http://terremotos.ing.uchile.cl/
Dias, A. M. P. G. (2012). Analysis of the nonlinear behavior of timber-concrete connections.

Journal of Structural Engineering (New York, N.Y.), 138(9), 1128; 1128-1137; 1137.
Dias, A. M. P. G., Van de Kuilen, J. W., Lopes, S., & Cruz, H. (2007). A non-linear 3D FEM
model to simulate timber–concrete joints. Advances in Engineering Software, 38(8–9), 522-
530. Retrieved on August 02, 2017 from
http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j.advengsoft.2006.08.024
120
Filliatrault, A., Tremblay, R., Christopoulos, C., Folz, B., & Pettinga, D. (2013). Elements of
earthquake engineering and structural dynamics (Third ed.) Presses Internationales
Polytechnique.
Fox, M., Sullivan, T., & Beyer, K. (2014). (2014). A case study in the capacity design of RC
coupled walls. Paper presented at the Second European Conference on Earthquake
Engineering and Seismology, Istanbul. 1-12.
FP Innovations. (2012). CLT handbook. FP Innovations, Canadian Edition
Fragiacomo, M. (2003). Design of bilinear hysteretic isolation systems. Earthquake Engineering

& Structural Dynamics, 32(9), 1333; 1333-1352; 1352.
Fragiacomo, M. (2011). Elastic and ductile design of multi-storey Crosslam massive wooden
buildings under seismic actions. Engineering Structures, 33(11), 3043; 3043-3053; 3053.
Gonzalez, M. (2017). Seismic response of tall buildings using ground motions based on national
building code of Canada 2015 (Master of Applied Science).
Government of British Columbia. (2009). Wood first act. BCLaws.
Green, D., Winandy, J., & Kretchmann, D. (1999). Mechanical properties of wood. Wood
handbook - wood as an engineering material (pp. 4-1). Madison, WI: U.S. Department of
Agriculture, Forest Service.
Guorui, H., & Toa, L. (2015). Review of baseline correction of strong-motion accelerogram.
International Journal of Science, Technology and Society, 3(6), 309-314.
Halchuk, S., Adams, J., & Allen, T. I. (2016). Fifth generation seismic hazard model for Canada:
Crustal, in-slab, and interface hazard values for southwestern Canada. Geological Survey of
Canada, Open File 8090, doi:1 .zip file. doi:10.4095/299244
He, M., Li, S., Guo, S., & Ni, C. (2011). The seismic performance in diaphragm plane of multi-
storey timber and concrete hybrid structure. Procedia Engineering, 14, 1606-1612. Retrieved
121
on November 15, 2017 from
http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j.proeng.2011.07.202
Hehl, S., Tannert, T., Meena, R., & Vallee, T. (2014). Experimental and numerical investigations
of groove connections for a novel timber-concrete-composite system. Journal of Performance
of Constructed Facilities, 28(6), A4014010; A4014010.
Ibarra, L. F., Medina, R. A., & Krawinkler, H. (2005). Hysteretic models that incorporate strength
and stiffness deterioration. Earthquake Engineering & Structural Dynamics, 34(12), 1489-
1511. doi:10.1002/eqe.495
Irvine, T. (2016). Damping properties of materials. Retrieved on June 06, 2017 from
www.vibrationdata.com
Isoda, H., Kawai, N., Koshihara, M., Araki, Y., & Tesfamariam, S. (2016). Timber-reinforced
concrete core hybrid system: Shake table experimental test. ASCE Journal of Structural
Engineering, doi:10.1061/(ASCE)ST.1943-541X.0001631, 04016152.
Isoda, H., & Tesfamariam, S. (2016). Connections for timber-concrete hybrid building:
Experimental and numerical model results. ASCE Journal of Performance of Constructed
Facilities, R1v2 doi:10.1061/(ASCE)CF.1943-5509.0000849, 04016024
Kalkan, E., & Chopra, A. (2010). Practical guidelines to select and scale earthquake records for
nonlinear response history analysis of structures. U.S. Department of the Interior U.S.
Geological Survey.
Khorsandnia, N. (2015). Coupled finite element-finite difference formulation for long-term

analysis of timber-concrete composite structures. Engineering Structures, 96, 139; 139-152;
152.
Klemenic, R., Fry, J. A., & Hooper, J. (2007). Performance-based design of tall reinforced concrete
ductile core wall systems. ASCE Structures Congress - Structural Engineering Research
Frontiers,
122
Kleven, M., & Noras, R. (2011). Finite element based simulation of diaphragm action on light
weight composite plywood to sheet metal roof element (Master).
Kremer, P., & Symmons, M. (2015). Mass timber construction as an alternative to concrete and
steel in the Australia building industry: A PESTEL evaluation of the potential. International
Wood Products Journal, 6(3), 138-147. doi:10.1179/2042645315Y.0000000010
Latour, M., & Rizzano, G. (2015). Cyclic behavior and modeling of a dissipative connector for
cross-laminated timber panel buildings. Journal of Earthquake Engineering, doi:
10.1080/13632469.2014.948645
Li, Z., He, M., Lam, F., Li, M., Ma, R., & Ma, Z. (2014). Finite element modeling and parametric
analysis of timber–steel hybrid structures. Structural Design of Tall and Special Buildings,
doi: 10.1002/tal.1107
Los Angeles Tall Buildings Structural Design Council. (2017). An alternative procedure for
seismic analysis and design of tall buildings located in the Los Angeles region: A consensus
document. Los Angeles: LATBSDC.
Lowes, L., Mitra, N., Altoontash. (2004). A beam-column joint model for simulating the
earthquake response of reinforced concrete frames. Pacific Earthquake Engineering Research
Center. Retrieved on October 16, 2017 from
http://peer.berkeley.edu/publications/peer_reports/reports_2003/0310.pdf
Lu, X., Cui, Y., Liu, J., & Gao, W. (2015). Shaking table test and numerical simulation of a 1/2-
scale self-centering reinforced concrete frame. Earthquake Engineering and Structural
Dynamics, doi: 10.1002/eqe
Meleki, H., Asiz, A., Smith, I., Gagnon, S., & Mohammad, M. (2011). Differential movements in
a timber multi-storey hybrid building. Procedia Engineering, 14, 1613-1620. Retrieved on
March 02, 2017 from http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j.proeng.2011.07.203
MGB Architecture, and Equilibrium Consulting. (2012). Tall wood.
123
Michaud, D., & Léger, P. (2014). Ground motions selection and scaling for nonlinear dynamic
analysis of structures located in eastern North America. Canadian Journal of Civil
Engineering, 41, 232–244. doi: dx.doi.org/10.1139/cjce-2012-0339
Monroy, M., Hull, A., & Atukorala, U. (2016). Technical memorandum: Massey tunnel
replacement project. (No. 14-1447-0012-008-TM-Rev0). Vancouver, BC: Golder Associates
Ltd.
Mori, T., Nakatani, M., & Tesfamariam, S. (2015). Performance of semi-rigid timber frame with
lagscrewbolt connections: Experimental, analytical, and numerical model results.
International Journal of Advanced Structural Engineering (IJASE), 7(4), 387-403.
doi:10.1007/s40091-015-0107-4
Mousseau, S., Paultre, P., & Mazars, J. (2008). Seismic performance of a full-scale, reinforced
high-performance concrete building. part II: Analytical study. Canadian Journal of Civil
Engineering, 35, 849-862. doi:10.1139/L08-019
Natural Resources Canada. (2015). Determine 2015 national building code of canada seismic
hazard values. Retrieved on June 12, 2017 from
http://www.earthquakescanada.nrcan.gc.ca/hazard-alea/interpolat/index_2015-en.php
Naumoski, N., Saatcioglu, M., Lin, L., & Amiri-Hormozaki, K. (2006). Evaluation of the effects
of spectrum compatible seismic excitations on the response of medium-height reinforced
concrete frame buildings. Canadian Journal of Civil Engineering, 33, 1304-1319.
NIED. (2018). K-net and KiK-net strong-motion seismograph network. Retrieved on February 22,
2018 from http://www.kyoshin.bosai.go.jp/kyoshin/quake/index_en.html
NRC - National Research Council Canada. (2010). National building code of Canada.
Panneton, M., Leger, P., & Tremblay, R. (2006). Inelastic analysis of a reinforced concrete shear
wall building according to the national building code of Canada 2005. Canadian Journal of
Civil Engineering, 33, 854-871.
124
Park, R., & Paulay, T. (1975). Reinforced concrete structures. USA: John Wiley and Sons.,
Paulay, T. (2001). Seismic response of structural walls: Recent developments. Canadian Journal
of Civil Engineering, 28, 922-937.
Paulay, T., & Preistley, N. (1992). Seismic design of reinforced concrete and masonry buildings
John Wiley and Sons, Inc.
Popovski, M., Mohammad, M., Ni, C., Rezai, M., Below, K., Malczyk, R., & Sherstobitoff, J.
(2014). Design and construction of tall wood buildings: Input data, testing and advanced
analysis.
Reyes, J., & Kalkan, E. (2012). (2012). Relevance of fault-normal/parallel and maximum direction
rotated ground motions on nonlinear behaviour of multi-story buildings. Paper presented at
the World Conference on Earthquake Engineering, Lisboa.
Robert Drysdale. (2014). In email correspondence to Chris Pieper (Ed.)
Rogers, G., Halchuk, S., Adams, J., & Allen, T. (2015). 5th generation (2015) seismic hazard
model for southwest British Columbia. Paper presented at the 11th Canadian Conference on
Earthquake Engineering Paper 94197,
Saaticioglu, M., & Humar, J. M. (2003). Dynamic analysis of buildings for earthquake-resistant
design. Canadian Journal of Civil Engineering, 338. doi:10.1139/L02-108
Schneider, J., Shen, Y., Stiemer, S. F., & Tesfamariam, S. (2015). Assessment and comparison of
experimental and numerical model studies of cross-laminated timber mechanical connections
under cyclic loading. Construction and Building Materials, 77, 197-212. Retrieved on July
22, 2017 from http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j.conbuildmat.2014.12.029
SeismoSoft. (2014). SeismoStruct user manual. SeismoStruct, Version 7.0
Shafiq, A. (2018). Performance-based seismic design and evaluation of eccentrically braced

frames with tubular links as bridge bents (Master of Applied Science).
125
Shen, Y., Schneider, J., Tesfamariam, S., Stiemer, S. F., & Mu, Z. (2013). Hysteresis behavior of
bracket connection in cross-laminated-timber shear walls. Construction and Building
Materials, 48, 980-991. Retrived on December 12, 2016 from
http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j. conbuildmat2013.07.050
Sherstobitoff, J. (2008). (2008). Time history versus response spectrum analysis. Paper presented
at the Time History Analysis Presented by CSCE, UBC Vancouver. 9-47.
Structurlam. (2016). Crosslam CLT technical design guide. Penticton, BC: Structurlam.
Tesfamariam, S., & Goda, K. (2017). Impact of earthquake types and aftershocks on loss
assessment of non-code conforming buildings: Case study with Victoria, British Columbia.
Earthquake Spectra, Preprint
Tremblay, R., Atkinson, G. M., Bouaanani, N., Daneshvar, P., Leger, P., & Koboevic, S.
(2015). Selection and scaling of ground motion time histories for seismic analysis using
NBCC 2015. Paper presented at the 11th Canadian Conference on Earthquake Engineering:
Canadian Association for Earthquake Engineering, Victoria, B.C. Canada.
Tremblay, R., Leger, P., & Tu, J. (2001). Inelastic seismic response of concrete shear walls
considering P-delta effects. Canadian Journal of Civil Engineering, 28, 640-655.
Ventura, C. (2005). Lecture 04 dynamic analysis of buildings. Unpublished manuscript.
Wallace, J. (2012). Behavior, design, and modeling of structural walls and coupling beams –
lessons from recent laboratory tests and earthquakes. International Journal of Concrete
Structures and Materials, Vol.6(No.1), 3-18. doi:10.1007/s40069-012-0001-4
Xu, B. H., Bouchaïr, A., Taazount, M., & Vega, E. J. (2009). Numerical and experimental analyses
of multiple-dowel steel-to-timber joints in tension perpendicular to grain. Engineering
Structures, 31(10), 2357-2367. Retrieved on July 18, 2017 from
http://dx.doi.org.ezproxy.library.ubc.ca/10.1016/j.engstruct.2009.05.013
126
Yasumura, M., Kobayashi, K., Okabe, M., Miyake, T., & Matsumoto, K. (2015). Full-scale tests
and numerical analysis of low-rise CLT structures under lateral loading. Journal of Structural
Engineering, E4015007-1 doi: 10.1061/(ASCE)ST.1943-541X.000
Yeoh, D. (2011). State of the art on timber-concrete composite structures: Literature review.
Journal of Structural Engineering (New York, N.Y.), 137(10), 1085; 1085-1095; 1095.
Zhang, C. (2015). Timber-concrete composite systems with ductile connections. Journal of

Structural Engineering (New York, N.Y.), 141(7)
Zhang, X., Fairhurst, M., & Tannert, T. (2015). Ductility estimation for a novel timber–steel hybrid
system. Journal of Structural Engineering, E4015001-2
127
Appendices
Appendix A – Additional information for selection of ground motions
Figure A-1 Deaggregation at 0.05 seconds – Significant contribution from crustal events
however in-slab dominating and interface is significant as well [(Canadian Hazards Information
Service, 2017) © with permission from publisher. This reproduction is a copy of an official work
that is published by the Government of Canada and the reproduction has not been produced in
affiliation with, or with the endorsement of the Government of Canada]
128
Figure A-2 Deaggregation at 0.2 seconds – In-slab dominates with crustal reducing in
importance and interface increasing [(Canadian Hazards Information Service, 2017) © with
permission from publisher. This reproduction is a copy of an official work that is published by
the Government of Canada and the reproduction has not been produced in affiliation with, or
with the endorsement of the Government of Canada]
129
Figure A-3 Deaggregation at 1.0 seconds – Interface events increasingly dominate [(Canadian
Hazards Information Service, 2017) © with permission from publisher. This reproduction is a
copy of an official work that is published by the Government of Canada and the reproduction has
not been produced in affiliation with, or with the endorsement of the Government of Canada]
130
Table A-1 Scenarios based on global mode and mean methods
Mean Mode*
Measure
Mw R (km) Mw R (km) Scenario
PGV 7.88 60 7.45 70 In-Slab
PGA 7.27 45 8.95 50 Interface
0.05 6.94 40 7.45 70 In-Slab
0.10 7.04 42 7.45 70 In-Slab
0.20 7.31 48 7.45 70 In-Slab
0.30 7.5 53 7.45 70 In-Slab
0.50 7.68 56 8.95 50 Interface
1.00 8.11 58 8.95 50 Interface
2.00 8.38 62 8.95 50 Interface
5.00 8.73 64 8.95 50 Interface
10.00 8.81 66 8.95 50 Interface
*Mode is sensitive to bin size but still a reasonable estimate. When
compared to Tremblay et al. (2015) it seems reasonable considering
distance between Vancouver and Victoria is about 90 km's and
somewhat orthogonal to the subduction zone fault. The mode is
certainly a better estimate than the mean which is clearly skewed by
multi-scenario contributions.
Table A-2 Scenarios based on inspection of deaggregation plots

Crustal In-slab Interface
Comments
M R M R M R
PGV 7.1 10 7.4 70 9 50 Target scenario's may be further refined by having access
PGA 6.8 10 7.4 70 9 50 to source specific deaggregation mode and mean values.
0.05 6.8 10 7.4 70 9 50
0.1 6.8 10 7.4 70 9 50 The global mode and means do not present the crustal
0.2 6.8 10 7.4 70 9 50 source as the most significant hazard at any period;
0.3 6.8 10 7.4 70 9 50 however, the crustal motions do contribute to the hazard
0.5 6.8 10 7.4 70 9 50 so a suite of crustal ground motions will be obtained.
1 6.9 10 7.4 70 9 50
2 7.0 20 7.4 70 9 50 Values within the period range of interest shown in bold.
5 7.3 20 7.4 60 9 50 Only these values contribute to the averages.
10 7.3 20 7.4 60 9 50
Target 6.8 10.0 7.4 70.0 9.0 50.0
131
Table A-3 Preliminary selection of three (3) suites of 11 ground motions
Crustal Target Scenario: 6.8 10 <1200
Record Database Earthquake Date Station Magnitude Hypocentral Soil (Vs30)
(dd/mm/yyyy) (Mw) distance (km) (m/s)
1 PEER-NGA-West2 El Mayor-Cucapah, Mexico 2010 El Centro Array #3 7.2 41 163
2 PEER-NGA-West2 Darfield, New Zealand 2010 Christchurch Resthaven 7.0 19 141
3 PEER-NGA-West2 Loma Prieta, USA 1989 Foster City - APEEL 1 6.9 44 116
4 PEER-NGA-West2 Iwate, Japan 2008 MYG006 6.9 30 147
5 PEER-NGA-West2 Northridge-01, USA 1994 Carson - Water St 6.7 50 161
6 PEER-NGA-West2 Tottori, Japan 2000 TTR008 6.6 7 139
7 PEER-NGA-West2 Superstition Hills-02, USA 1987 Imperial Valley WLA 6.5 24 179
8 PEER-NGA-West2 Imperial Valley-06, USA 1979 El Centro Array #3 6.5 13 163
9 PEER-NGA-West2 Christchurch, New Zealand 2011 Christchurch Resthaven 6.2 5 141
10 PEER-NGA-West2 Parkfield-02, USA 2004 Parkfield - Fault Zone 1 6.0 3 178
11 KiK-Net/K-Net Fukushima-Hamadori, Japan 11/4/2011 IBRH10 6.7 11 144
In-slab Target Scenario: 7.4 70 <200
1 University of Chile / u.chile Tarapaca, Chile 13/06/2005 TARA05R (Pica) 7.8 125 492
2 University of Chile / u.chile Tarapaca, Chile 13/06/2005 TARA09R (IDEM) 7.8 144 386
3 Seismotoolbox.ca Stochastic, Atkinson 2005 r20/04/2018 M7wnaE1_1_2 7.5 70* 180
4 Seismotoolbox.ca Stochastic, Atkinson 2005 r20/04/2018 M7wnaE1_3_4 7.5 70* 180
5 KiK-Net/K-Net Tohoku-Aftershock, Japan 11/3/2011 AOMH13 7.4 144 154
6 University of Chile / u.chile Punitaqui, Chile 14/10/1997 Illapel 7.1 112 502
7 KiK-Net/K-Net Hokkaido Aftershock_1, Japan 6/12/2004 NMRH04 6.9 77 168
8 KiK-Net/K-Net Obihiro, Japan 2/2/2013 TKCH07 6.9 107 140
9 PNSN / COSMOS VDC Nisqually, Olympia, USA 28/02/2001 DNR Bldg / near 7054 6.8 56 300
10 PNSN / COSMOS VDC Nisqually, Olympia, USA 28/02/2001 Crown Plaza/ near 7040 6.8 78 285
11 KiK-Net/K-Net 2001 Geiyo, Japan 24/03/2001 EHMH04 6.7 65 254
Interface Target Scenario: 9 50 <200
1 KiK-Net/K-Net Tohoku, Japan 11/3/2011 IBRH10 9.0 338 144
2 KiK-Net/K-Net Tohoku, Japan 11/3/2011 IBRH07 9.0 328 107
3 Seismotoolbox.ca Stochastic, Atkinson 2005 r20/04/2018 M9wnaE1_1 9.0 60 180
4 Seismotoolbox.ca Stochastic, Atkinson & Macias 2008 r20/04/2018 amp-vic_1 9.0 60 760
5 University of Chile / u.chile Maule, Chile 2/27/2010 CONC 8.8 69 223
6 University of Chile / u.chile Maule, Chile 2/27/2010 CCSP 8.8 70 236
7 Seismotoolbox.ca Modified Tokachi, Atkinson & Macias r20/04/2018 vic_084ew 8.5 72 760
2008
8 KiK-Net/K-Net Kuril, Japan 15/11/2015 NMRH04 8.4 790 180
9 University of Chile / COSMOS Valparaiso, Chile 3/3/1985 RANC01S (Rapel) 8.0 85 108
VDC
10 KiK-Net/K-Net Hokkaido (Tokachi-Oki), Japan 26/09/2003 TKCH07 8.0 125 140
11 KiK-Net/K-Net Hokkaido (Tokachi-Oki), Japan 26/09/2003 KSRH02 8.0 150 219
132
Table A-4 Detailed notes for final ground motion suites (continues horizontally on next page)
Basic identification Pre-processing Target for selection
Time
Time
Hazard Date step Baseline Filter Magnitu Magnit Distance Soil
step D,o Distanc Soil target
and Earthquake (DD/MM/ Inter- correction applie de target ude target site
initial (s) e (km) (m/s)
range YYYY) polated applied d (Mw) (Mw) (km) (m/s)
(s)
(s)
149.9
H1 September 4, 95
Darfield, NZ Quadratic 7.0 19
2010 149.9
141
H2 80
Christchurch, H1 February 22, 0.005 Pre- 50.00
6.2 5
NZ H2 2011 proces 0
Crustal Linear
Imperial H1 sed by
0.03 < October 15, 39.59
Valley, CA, others 6.8 6.5 10 13 163
Trs < 1979 0
USA H2 (no
0.40
update 299.9
H1 October 6, req'd.) 70
Tottori, Japan 0.010 Cubic 6.6 7 139
2010 299.9
H2 60
Parkfield, CA, H1 September 28, 20.97
0.005 Linear 6.0 3 178
USA H2 2004 0
Pica - H1 Butter 251.9 Site class E
125 492
Tarapaca, Pica - H2 June 13, worth 80 (V<180m/s)
0.005 Quadratic 7.8 but
Chile Idiem - H1 2005 (0.1 - 55.80
0.010 144 practically 386
Idiem - H2 25 Hz 0
In-slab Bandp 100<V<200
H1 February 2, 299.9
0.03 < Obihiro, Japan 0.010 Linear ass) 6.9 107 is a realistic 140
H2 2013 80 7.4 70
Trs < target range
0.80 no update
95.98
Crown - H1 req’d. no 78 285
Nisqually, February 28, 0
Crown - H2 0.005 Linear update 6.8
WA, USA 2001
DNR - H1 no update req'd. 151.9
56 300
DNR - H2 req’d. 80
H1 March 11, 262.4
Tohoku, Japan 0.010 9.0 338 144
H2 2011 80
Quadratic
Conc - H1 Butter 141.6
69 223
Interface Conc - H2 February 27, worth 70
Maule, Chile 8.8
0.30 < CCSP - H1 2010 (0.1 - 143.2
Cubic 9.0 50 70 236
Trs < CCSP - H2 25 Hz 60
1.50 0.005 Bandp
T - H1
ass) 125 140
Hokkaido, T - H2 September 26, 299.9
Linear 8.0
Japan K - H1 2003 90
150 219
K - H2
133
Table A-4 Detailed notes for final ground motion suites (concluded)
Basic identification Linear scaling results Detailed identification
Suit
Hazard Date Indivi e Data
Record Data Flatfile/docu
and Earthquake (DD/MM/ dual scale Comments Station and Channel retrie
ing site base mentation
range YYYY) scale fact val
factor or
H1 RSN6959_DARFIELD_REHSN02E
Darfield, NZ September 4, 2010 1.64
H2 RSN6959_DARFIELD_REHSS88E Restha
Christchurch, H1 Balance RSN8123_CCHURCH_REHSN02E ven
February 22, 2011 1.38
NZ H2 minimizing RSN8123_CCHURCH_REHSS88E
Crustal
Imperial H1 scale factor RSN178_IMPVALL.H_H-E03140
0.03 < E.C.A
Valley, CA, October 15, 1979 1.69 1.04 with M-R PEER NGA West2
Trs < RSN178_IMPVALL.H_H-E03230 #3
USA H2 relationship
0.40
H1 s and soil RSN3965_TOTTORI_TTR008EW TTR00
Tottori, Japan October 6, 2010 1.48 conditions
H2 RSN3965_TOTTORI_TTR008NS 8
Parkfield, CA, H1 September 28, RSN4107_PARK2004_COW090 Parkfie
0.88
USA H2 2004 RSN4107_PARK2004_COW360 ld FZ 1
Pica - H1 PICA S/N 2799 - CHAN 1: EW
0.88 Pica
Tarapaca, Pica - H2 Relaxed PICA S/N 2799 - CHAN 2: NS Universidad de Bastias et al
June 13, 2005
Chile Idiem - H1 soil IQUIQUE IDIEM S/N 7051 - CHAN 1: EW Iquique Chile - Renadic (2016)
2.61
Idiem - H2 matching IQUIQUE IDIEM S/N 7051 - CHAN NS Idiem
In-slab criteria;
H1 TKCH071302022317 - EW TKCH Dawood et al
0.03 < Obihiro, Japan February 2, 2013 1.53 K-Net/KiK-Net
H2 1.63 TKCH071302022317 - NS 07 (2016)
Trs <
0.80 Crown - H1 7010g Seattle
2.07 Crown Filliatrault
USGS COS
Nisqually, Crown - H2 Not free- 77010j Plaza et al (2011);
February 28, 2001 / MOS
WA, USA field ESCWEB
DNR - H1 7015a PNSN VDC
2.64 DNR (r2018)
DNR - H2 7015c
H1 IBRH101103111446 - EW IBRH1 Dawood et al
Tohoku, Japan March 11, 2011 1.63 K-Net/KiK-Net
H2 IBRH101103111446 - NS 0 (2016)
Conc - H1 Concepcion1002271 S/N 5003 - CHAN 1 L
1.75 Conc
Conc - H2 Concepcion1002271 S/N 5003 - CHAN 2
Interface Universidad de Bastias et al
Maule, Chile February 27, 2010 Constitucion1002271 - S/N 4598 - CHAN 01
0.30 < Relax soil Chile - Renadic (2016)
CCSP - H1 1.8 1.24 L Cons
Trs < criteria
CCSP - H2 Constitucion1002271 - S/N 4598 - CHAN 2
1.50
T - H1 TKCH070309260450 - EW TKCH
1.05
Hokkaido, T - H2 September 26, TKCH070309260450 - NS 07 Dawood et al
K-Net/KiK-Net
Japan K - H1 2003 KSRH020309260450 - EW KSRH (2016)
1.37
K - H2 KSRH020309260450 - NS 02
134

Seismic Analysis and Design of Hybrid Concrete Timber Structures With 2015 National Building Code of

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Seismic Analysis and Design of Hybrid Concrete Timber Structures With 2015 National Building Code of

Uploaded by

Copyright:

Available Formats

SEISMIC ANALYSIS AND DESIGN OF HYBRID

CONCRETE TIMBER STRUCTURES WITH 2015

Christopher Gregory Pieper

B.A.Sc., The University of British Columbia, 2014

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF APPLIED SCIENCE

THE COLLEGE OF GRADUATE STUDIES

THE UNIVERSITY OF BRITISH COLUMBIA

© Christopher Gregory Pieper, 2018

SEISMIC ANALYSIS AND DESIGN OF HYBRID CONCRETE TIMBER

submitted by Christopher Gregory Pieper in partial fulfillment of the requirements of

the degree of Master of Applied Science.

Dr. Solomon Tesfamariam, School of Engineering (Civil)

Dr. Shahria Alam, School of Engineering (Civil)

Dr. Kasun Hewage, School of Engineering (Civil)

Dr. Thomas Johnson, School of Engineering (Electrical)

Particularly in moderate to high seismic regions, earthquake engineering is required to ensure

Abstract .......................................................................................................................................... iii

Lay Summary ................................................................................................................................. iv

Table of Contents ........................................................................................................................... vi

List of Tables .................................................................................................................................. x

List of Figures ................................................................................................................................ xi

Acknowledgements ...................................................................................................................... xiv

2 Literature Review .................................................................................................................... 7

4 Case Details ........................................................................................................................... 37

7 Conclusions ......................................................................................................................... 109

Appendices .................................................................................................................................. 128

Table 4-1 Structural loads and deflection criteria ........................................................................ 49

Table 5-2 Geotechnical parameters provided for the site ............................................................ 59

Table 5-3 NBCC (2015) Table 4.1.8.4.-B ................................................................................... 61

Table 5-4 Structural loads and deflection criteria ........................................................................ 63

Table 5-5 Timber component sizes .............................................................................................. 64

Table 5-6 Preliminary seismic weights used for analysis ............................................................ 66

Table 5-7 Eccentricities and total static torsional moment .......................................................... 69

Table 5-8 Change in key parameters due to model update .......................................................... 72

Table 5-10 Participating mass ratio's ........................................................................................... 78

Table 5-11 Target scenario's ........................................................................................................ 80

Table 5-12 Scenario-specific period range of interest ................................................................. 80

Table 6-2 Pivot parameters for lagscrewbolts connections .......................................................... 93

Table 6-3 Final selected suites of ground motions ..................................................................... 101

Figure 1-1 12-Storey concrete timber structure ............................................................................. 1

Figure 2-1 Hysteretic connection models .................................................................................... 23

Figure 4-1 Experimental apparatus for connection testing .......................................................... 37

Figure 4-2 Load protocol ............................................................................................................. 38

Figure 4-3 Connections ............................................................................................................... 39

Figure 4-4 Portal frame study ...................................................................................................... 39

Figure 4-5 Portal frame lagscrewbolt connections ...................................................................... 40

Figure 4-6 General arrangement of concrete-timber hybrid ........................................................ 41

Figure 4-7 Three-dimensional two-storey concrete-timber structure .......................................... 42

Figure 4-8 Concrete core section ................................................................................................. 43

Figure 4-9 General arrangement plans ......................................................................................... 45

Figure 4-11 Core walls ................................................................................................................. 47

Figure 4-12 Preliminary load criteria ........................................................................................... 48

Figure 4-13 Preliminary structural model .................................................................................... 49

Figure 5-1 One-dimensional model with hysteretic spring .......................................................... 50

Figure 5-2 Two-dimensional non-linear portal frame model....................................................... 54

Figure 5-3 Hysteretic matching of column base in frame model ................................................. 55

Figure 5-6 Design spectral response accelerations ...................................................................... 62

Figure 5-7 Preliminary shear and overturning results .................................................................. 69

Figure 5-8 Varying eccentricities and inherent torsion ................................................................ 75

Figure 5-9 Scaling of elastic forces.............................................................................................. 75

Figure 5-10 Scaling of drifts considering torsion ........................................................................ 76